diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubes / Valves (http://www.diyaudio.com/forums/tubes-valves/)
-   -   How to get the best out of the 6C33C-B? (http://www.diyaudio.com/forums/tubes-valves/170506-how-get-best-out-6c33c-b.html)

knightsound 20th July 2010 05:10 AM

How to get the best out of the 6C33C-B?
 
1 Attachment(s)
The 6C33C-B is an attractive tube; 15-20W output, relatively inexpensive at $16 each in Russia and, at 200V, relatively low voltage. The downside? The 6C33C-B requires a relatively high current, typically 220mA at 200V. And it has stinkin' hot heaters.

Many of the published 6C33C-B circuits use variable bias, and globs of feedback. Okay in itself, but I have a feeling it's possible to do better, possibly with CCS's and a "fixed" bias approach.

In the search for the best tube performance, constant current supply techniques seem to be very popular. For example, those used in the Baby Huey, Tubelab SE, and SY's Red Light District, attract my attention. Which technique would be most appropriate for the high current 6C33C-B?

The approach I'm favouring is George Anderson's PowerDrive (PowerDrive Cookbook), as featured in the Tubelab SE.

With the substitution of the Fairchild FQP4P40 instead of the IXYS 10M45, PowerDrive should work, but is it the best idea? Has anyone tried this before? Or, is there a different approach I've missed, which may be better in terms of a clean, linear current delivery?

My design goals are straightforward:
- One 6C33C-B per channel, single-ended operation
- 15-20W @ < 1% distortion
- Fixed bias
- Thermal efficiency, where possible (ie. SMPS in PS).
- Circuit designed to safely accommodate fluctuations in mains supply voltage. This is important. Frequent power outages, and variations that cause the lights to dim, necessitate a fail-safe circuit design.

I have three questions:

- Is PowerDrive the best option for this application?

- Would a choke and a MOSFET follower be worthwhile on the input tube?

- I plan to use a grid stopper resistor on the output tube. For output stage biasing, what would be best? A cascade CCS, or perhaps a variation on SY's cunning, thermally efficient LED approach (how many bulbs?!! I'm not sure if this would be feasible, but I wonder if there's an alternative to LED's).

For the input stage my choice is limited to what I have on hand, either/or: 6N1P-EV (=E88CC = 6DJ8 = 6922); 6N2P-ER (=12AX7 = ECC83); 6N6P (=ECC99 = E182CC). I am thinking of a simple two stage 6N1P-EV -> 6C33C-B. I'm wary of additional complication in a three stage circuit design. Bear in mind that by using a CCS, the input stage gain will increase, but to estimate by how much, can anyone suggest a suitable calculation?

Oh dear, that's more than three questions. :)

mace1337 20th July 2010 09:44 AM

This thread could be very interesting for me, since I'm in the same boat as you basically.

I have 4 6C33C-B tubes, two 5C8S rectifiers (I just thought they looked cool together :D ) but a bit larger choice of driver tubes. (6N1P, 6N2P, 6N3P, 6N6P, 6N23P (the reflektor 6922 was based on this, 6922EH among others) and some more I can't remember, but all USSR stuff).

I have seen some interesting designs using Russian 6E5P and 6E6P tetrode tubes as drivers, check out the GoodSoundClub - Romy the Cat's Audio Site forums for lots of info on 6C33C amplifier design, the survival guide is a good place to start.

knightsound 21st July 2010 05:34 AM

mace1337, thanks for the links. Glad to have you aboard! Is it okay if we start with just one 6C33C-B for output? I need to start with the basics, then you can see where I have screwed up. :)

I forgot to mention, a first pass of the BOM for this amp works out less than $750, depending on OPT choice. More on that later.

Here's a first draft of the circuit:

http://www.lightspell.com/images/kni...6C33CBv0.1.jpg

6C33C-B ($16 from rutubes.com) - The output tube will need a bias current somewhere between -55V to -67V. (derived from two different sets of curves. Does anyone have some of their own measured values?)

The 6N2P-EV ($2.20-$3.80 from rutubes.com) - Note, similar to, but not the same as, the 12AX7. The filaments, pins 4 and 5, are wired in parallel which means they must have 6.3V, 300mA supply (not 12.6V). Pin 9 should ground to the chassis.

mu = 97.5
Ri 50 KOhm
Transconductance 2.25mA/V
Plate Voltage 300V
Plate Dissipation 1W per triode
Cathode Current 10mA per triode
Rgrid 0.5M
Cin 2.7pF
Cout 2.9pF/3.1pF for each triode

I confess I'm not sure how to work out the component values on the right hand side where the 2SK2700 joins the circuit. I'll start with the MOSFET spec. sheet, but does anyone have any clues?

mace1337 21st July 2010 08:42 AM

I was thinking on going PP or maybe PSE but because I am also rather inexperienced with amplifier design I think a simple SE amp would be best to start with.

That being said, since the 6C33C is internally also two triodes it may be relatively easy to parallel two tubes (so we have 4 triodes in parallel).

This thread will hopefully be a nice learning experience for me. :D

WT 21st July 2010 09:08 AM

6c33 is a single triode with 2 separate filaments. You can only use it as single triode. If you use one filament you can use it up to 45 watts, with 2 filaments used up to 60 watts.

Knightsound, your schematic show 6c33 as quadtrode. I guess you misunderstood or drawing mistake. I don't think your schematic will work. Your mosfet current source will not work as you draw.

I build 6C33 similar to Simplex amp many years ago. First, I build as the article but later I change the 6sn7 to penthode drive which I like better. The amp sound very good. I bias 6c33 at 225V 200mA. This amp run hot.

This is the simplex article. If you are first time builder, I am not recommend you to build it.
http://www.polisois-audio.com/docume...20Redux%20.pdf

mace1337 21st July 2010 09:32 AM

Quote:

Originally Posted by WT (Post 2249424)
6c33 is a single triode with 2 separate filaments. You can only use it as single triode. If you use one filament you can use it up to 45 watts, with 2 filaments used up to 60 watts.

As I understand it, there are two triodes in there, the anode, cathode and grid are separate, but connected in parallel internally, so they act as one.

Only the filament is separate, but this enables you to use just one of the two triodes.

ilimzn 21st July 2010 11:31 AM

Quote:

Originally Posted by knightsound (Post 2248442)
The 6C33C-B is an attractive tube...
...
My design goals are straightforward:
- One 6C33C-B per channel, single-ended operation
- 15-20W @ < 1% distortion
- Fixed bias
- Thermal efficiency, where possible (ie. SMPS in PS).
- Circuit designed to safely accommodate fluctuations in mains supply voltage. This is important. Frequent power outages, and variations that cause the lights to dim, necessitate a fail-safe circuit design.

I am sorry to point out that your questions are largely irrelevant before you do the required basic 'legwork' around the 6S33S (this would be the latin transliteration, C is cyrillic for S), and in basic circuit theory - as is evidenced by your diagram. The 6S3S is a triode, not tetrode, Your output stage only has ground, no power supply. There are a few subtleties with the MOSFET follower which you seem to not understand (deriving of bias voltage from the negative pulldown voltage for the MOSFET)... but, most of all, your 'straightforward' design goals are not only not straightforward, but impossible without a lot of feedback.

I would suggest you start with reading and understanding the 6S33S datasheet. Just because it's said you can get 15-20W out of it, it does not mean you can have that at any distortion figure you choose. Indeed, you would be hard pressed to get those 15-20W while having the tubes last!
Start with the plate curves for the 6S33S. Note that using both cathodes at a plate voltage over 250V severely derates available dissipation and thus output power. Note that the datasheet gives a MAXIMUM grid voltage of -0.5V, so the tube is nnot designed for positive grid operation, and furthermore, has substantial grid current close to Vg=0V, which means that a follower might be handy in the driver circuits, but it's output must be limnited not to overdrive the grid. Learn how to plot a load line across plate curves to estimate power output and distortion, a search of this forum should give you lots of pointers. Before you do these basics, discussing finesses of driver and follower circuits are quite irrelevant.

ilimzn 21st July 2010 11:37 AM

Quote:

Originally Posted by mace1337 (Post 2249414)
I was thinking on going PP or maybe PSE but because I am also rather inexperienced with amplifier design I think a simple SE amp would be best to start with.

That being said, since the 6C33C is internally also two triodes it may be relatively easy to parallel two tubes (so we have 4 triodes in parallel).

This thread will hopefully be a nice learning experience for me. :D

It would be a learning experience with a lot of smoke, I am afraid. Even a short search of the forum will turn up that there are indeed LOTS of problems simplly paralelling 6S33S, as indeed happens with all high mutual conductance tubes. The 6S33S is particularly troublesome because there are great variations from tube to tube. Indeed, it's not advisable to use it in fixed bias mode unless at least some of the bias voltage is derived from a cathode resistor - in case of parallelling, each tube must have it's own, or at least it's own bias network, so balance can be set. It is ultimately possible to run them this way but only after burn-in during which biac surrent will be very unstable.

The tube itself has two cathodes and grids but only a single massive plate. Even so, when both cathodes are running, you DON'T get twice the capability, indeed, you only get 30% more, this is to account for the massive amount of heat generated by the heaters themselves, as well as electrode system imbalance. It's all there in the datasheet!

mace1337 21st July 2010 12:27 PM

Thanks for your comments. :)

I know about the cyrillic, the strange thing about this tube is that it's generally referred to as the 6C33C-B instead of the more correct (depending on your viewpoint of what is "correct") 6S33S-V.
It's moreso strange because other Russian tubes do get translated most of the time.

Anyway, I intend to take it slow and learn about designing amps, flesh out the design and in a few weeks or months time start building. I am in no kind of rush.

This will not be my first amp build or tube amp build, I have made a few before, but never from my own design, or with such an interesting and "difficult" tube.

Right now I have a few PDF's about designing and building tube amps and I'm reading them with great interest.

djn 21st July 2010 02:47 PM

After looking at all the schems I could find for the 6C33C, I landed on Tim Mellow's OTL. Unfortunatly I can't attache the write up on the amp because it is too big (file wise) I will PM it to you. Very simple build and no OPT so it is cheaper to build than others. I have the PT already and a handfull of misc. parts so just need to order caps and res.

Cheers.


All times are GMT. The time now is 03:00 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2