| Search | Login/Register
   Home » Melquiades Amplifier » To drive the 6C33C... (66 posts, 4 pages)
  Print Thread | 1st Post |  
Page 2 of 3 (66 items) Select Pages:  « 1 2 3 »
   Target    Threads for related reading   Most recent post in related threads   Forum  Replies   Views   Started 
  »  New  6C33C myths: audio Moronometr...  Overdrive warning light...  Melquiades Amplifier  Forum     3  56367  06-22-2005
  »  New  The short "6C33C Survival Guide"...  Ac filament.....  Melquiades Amplifier  Forum     20  363209  12-18-2007
  »  New  A different breed of 6C33C amplifier...  6C33C mono block @ clipping...  Melquiades Amplifier  Forum     84  869110  09-06-2011
  »  New  6C33C socket types, where to get the ideal one?..  Not so much the material that matters to me but the pin...  Melquiades Amplifier  Forum     2  29463  01-09-2012
03-22-2006 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 26
Post ID: 2233
Reply to: 2230
Re: lips raning and Lamm

 guy sergeant wrote:
I have heard Lamm ML2's in a couple of different people's home systems. Did they know how to use them properly? Well either they didn't (quite possible) or the amplifiers simply lacked the virtues I would prefer.

Well, I never heard any credible and rational critics of Lamm ML2 performance from anyone and I know quite a lot of ML2 users. Since I publish my personal unsatisfactory comments about ML2 I begin to hear here and there that some people do not like that ML2’s Sound or some of the fools have it for sale. I always follow up those comments and in some cases I visited the people and familiarized myself the context of the actual results that people achieved in their rooms before the Lamm ML2 was “not good enough”. What I have witnessed that NONE of them even remotely reach the point where was reasonably to talk about the Lamm’s limitations. People juts ran their lips, nothing else…

 guy sergeant wrote:
Tommy Horning is the guy who makes those speakers but he wasn't using Lamms. His speakers are very enjoyable though.

It would be fun to combine a highpassed at 500Hz Hornins as MF channel with something that would be real bass and real upperbass capable, without stressing that flimsy rice-paper Hornings’ driver. The Hornings’ MF is very interesting though….


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
03-22-2006 Post does not mapped to Knowledge Tree
Genn
Moscow
Posts 16
Joined on 06-28-2005

Post #: 27
Post ID: 2236
Reply to: 2218
Re: Some spaghetti of thoughts: load and capacitors
I have read once again the explanation of the Melq's project and find 2 interesting things:
- load for full range amp should be 1200 Ohm - that is very good for the tube and provides good damping,
- capacitors in case of LF channel - 20000 uF (!).
The last value is (for me) a good proof of bass quality, and overall quality as well.
03-22-2006 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 28
Post ID: 2237
Reply to: 2236
A simple SET isn't good amplifier….

I do not really believe in the ultimate ability of SET amplifiers to drive loudspeakers, at least any loudspeakers. I am not talking about the impedance fluctuations or the inappropriately low impedance dives. Those are well know issues. The people who use 4R loudspeakers, that dive to 2.3R, drive them with SETs and then bitch that “the amp has no bass” are juts the Morons. There is another big problem with SETs that from my point of view make them not suitable for full range operation.

 Genn wrote:
load for full range amp should be 1200 Ohm - that is very good for the tube and provides good damping.

The good damping for what? It is very very very frequently, practically always, that the MF and LF drivers (combined with the way how they implemented) would require different loading of the output tube. In my case the 6C33C loaded to 1200 Ohm and it is fine for my MF channels but I drive 6C33C much harder for LF channels (approximately 700R). It is not necessary has a lot of to do with 6C33C but with many other parameters that are external to amplifier. (For instant my LF speakers are superbly sensitive for damping and 50R of plate loading is quite auditable)

The point is that when people use SET for a full range then the by default introduce a huge compromise because the LF and MD simultaneously load the plate of the output stage. This is why I do not really believe that a SET might drive a full range inhalation. Usually SET does fine above 60Hz (with minimum capacitance in PS, and no compromise in transformer to accommodate the LF) and to go lower it needs another SET optimized for LF (with max capacitance in PC, oversized inductance on primary and exact LF loading of anode). This is where my rule of DSET (dedicated SET) was bourn…. A simple SET is just a MF amplifier….

The caT


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
03-22-2006 Post does not mapped to Knowledge Tree
Genn
Moscow
Posts 16
Joined on 06-28-2005

Post #: 29
Post ID: 2238
Reply to: 2237
Re: A simple SET isn't good amplifier….
 Romy the Cat wrote:
I do not really believe in the ultimate ability of SET amplifiers to drive loudspeakers, at least any loudspeakers.

Well - I've heard a couple of such amplifiers, made by Mr.Juri Makarov.  They did have more or less similar design concept: 3x6S19P @ 3.5k load, specially designed OPT, 5W, 2.1 .. 2.5 Hz to 200 kHz, about 0,5 Farad total capacitance, 2.1 .. 2.5 Hz to 200 kHz, direct coupled driver on 6S45P, 70..100 kg per monoblock.   First used regulated power supply on Gu-50, and stbilised filament, the last was most simple and did use only stabilised filament. 

The Designer used a special listening room with lot of natural pine wood deflectors in it. It was very quiet.
Regular loudspeakers was used there - initially top from Mirage family, discontinued probably 10 years ago, now - approximatly the top speakers from Montana brand.

Well, he is not respected by the Morons.

p.s. An important thing of this amp construction was the cooling concept. All tubes were hidden in the specialy milled radiator - to provide a good temperature stability for the tubes, and according to the words of designer - amplifier was switched off only several times a year. 

Romy, looking through some suppliers I've find out Pearl coolers for 6S33S. May be it is worth to give them a try, since the high temperature consequences is an issue with 6S33S?
03-22-2006 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 30
Post ID: 2239
Reply to: 2238
Cooling the 6C33C.

It is completely irrelevant how frequently an amp switched off and on if a soft start is made properly. A few years ago my friend and I, having the 6C33C based Lamms ML2 (where I do not like how the soft was working) , noted that we both change the output tubes once a year despite that I turned it on and of dally but the friend of mine ran it all year long, practically without shutting it down.

Cooling the 6C33C? The original Melquiades was running with natural convention and it was fine. The Super Melquiades quite packed inside and with all that it has in this chassis it getting too hot. The Super Melq uses a fan that outputs approximately 75% of its flow to the LF tube. It is, 1300 rotations @12DC with 26dB fan sitting at 9V. The fan is well decoupled from the chassis but still with a stethoscope it is auditable on the chassis. The way in which this fan works in the Super Melq’s enclosure make me happy, I initially though to put a thermostat in there but then realized that it is good keep running the fan all time.

The use external leaves on 6C33C is kind of completed. The 6C33C is not flat and has two diameters. It is not clean what diameters should be cooled, even with a forced convention. The Pearl coolers are fine but Bill’s cooler are half ass coolers and the type of the coolers that Pearl has very limited thermal conductively between the glass and the cooler. I had an idea a few years ago to press the 6C33C into a form and after it solidify to fill it with copper. So, what I would have is a flat contact (without the “nipples” that should be omitted and enlarged) between a copper sarcophagus and the glass. Then too the sarcophagus it might be attached any conventional cooling devise. This head-sarcophagus might be completely removable and might just “condom” the 6C33C.  When I designed the Super Melq holes around the 6C33C I thought that I might try to do it sometimes. I think it will be $300-$400 per single cooler… kind of pricey and not really necessary….

The Cat


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
03-22-2006 Post does not mapped to Knowledge Tree
hagtech


Hawaii
Posts 117
Joined on 02-13-2006

Post #: 31
Post ID: 2240
Reply to: 2239
Re: Cracking the 6C33C.
I think you'll have trouble adding a copper sleeve to the power tube.  The thermal coefficient of expansion for copper is probably much greater than that of the glass used.  Hence, the copper will enlarge much faster and break the glass.  It's a good idea, just the wrong material.  Obviously you are familiar with the water cooled transmitting tubes.  Your molded sleeve concept can probably work with the right material. 

On the other hand, most of the heat from a tube is radiated, not conducted.  Therefore, cooling the glass won't help too much.  You have to cool the electrodes (plate) inside.  Really big power tube give you metal contact with the electrodes (forget the glass).  There is little thermal conductivity between the glass and metal, especially with a vacuum in the middle.  What you don't want to do is surround the tube with mirrors (which a number of amplifier makers do).  The mirrors will reflect the radiated heat right back to the tube, forming an oven.  Black surfaces (or none at all) work best.

If you use a fan, aim it at the base of the tube.  That is where you need the convection or airflow.  Also, make sure the tube isn't radiating all that power into other chassis components.

As far as soft start, I believe that is where your choke rectified heaters come in.  Such an arrangement is extremely gentle.

jh
03-22-2006 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 32
Post ID: 2241
Reply to: 2240
Re: Cooking the 6C33C.

Yes, Jim, you put it very well. Still in case of the 6C33C it only possible to cool the glass and I made experiments blowing a strong flow to the glass and it did cool down the plates. The problem is that it is not a good idea to apply a strong flow juts to one side of the tube and the entire balloon should be cooled down. To do it with sufficient flow it would need a very powerful blower and consequentially a lot of noise. It would be possible to engineer something effective for the 6C33C cooling but the real question that I always ask myself: why would I need it? It the end, leaving aside my foolish desire to deal with imaginary problems, the 6C33C does not need cooling…. In the case of the 3 tubes in a relatively small chassis (the Super Melq’s gain chassis is 14x20), and considering that the chassis is staffed the forced airflow to the tubes socket is really necessary, at least in my case.

 hagtech wrote:
  As far as soft start, I believe that is where your choke rectified heaters come in. Such an arrangement is extremely gentle.
Actually I use choke-rectified heaters in my phonostage not in the power amps. Melquiades has both stages filaments powered by AC. The 6C33C filaments never fail and the only thing that should be a reason of concern with 6C33C is its anode. Actually it is funny but I know a guy who use 6C33C at barbarian power and fixed bias (good for sound), So, he monitors the plate temperature and if it goes up then he lower voltage on the filaments…. He is loosing power this way but… hey… whatever makes him happy… So in my case I soft-start the plate supply. The filaments do off. In 100 sec I apply bias voltage and the voltage on the B+ of the first tube. In 20 sec I turn on B+ on the 6C33C. The voltage begin to charge the last large cap for ~3 sec and then reaches the cruse voltage of the 230V. If the nominal plate current is 200mA then it reaches approximately 110mA after the voltage is set. Then within next 10 minutes the current gradually rises to 200mA and stay there forever.

In the end… all of this is not really important as getter on the 6C33C burns out sooner then any other danged take palace on this tube…. Pretty much a year of operation and you need to pay another $10 for a new 6C33C....

Rgs,
Romy the Cat


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
03-23-2006 Post does not mapped to Knowledge Tree
Genn
Moscow
Posts 16
Joined on 06-28-2005

Post #: 33
Post ID: 2242
Reply to: 2241
Saving the 6C33C.
Well - there is a technical literature related the cooling of the tube.  According to it the temperature of the glass surface is not equall - it is hire in the places, where heater is visible.  So - you can make a pair of radiators, attache to the tube on both sides and extend it's lifetime.  Another issue of the 6C33C - short lifetime of the socket.  I think, that ventilation will help to extend their lifetime as well.
03-23-2006 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 34
Post ID: 2243
Reply to: 2242
6C33C: don't use badly-made sockets.

 Genn wrote:
Another issue of the 6C33C - short lifetime of the socket.  I think, that ventilation will help to extend their lifetime as well.

It is not a problem with the 6C33C but with the stupidity of the very specific Russian people who designed the sockets for 6C33C. Do not use the Russian or Chinese made socket and you will not have the “issues”. The Russian and Chinese sockets form a triangle to hold the 6C33C pins. How stupid more it should be if the contact between the socket's pins and the tube's pins is basically a single point, as the socket’s “leaves” that accept the tube pins, open up to become “parallel” to the pin but in the reality it never parallel?

Johnson’s sockets for 6C33C (it was Johnson’s $236 or Johnson’s #237 if I am not mistaken) have no problems to hold the 6C33C and they do not burn itself. They have a contact surface as a short cylinder instead of triangle and when the “leaves” open up the pins are hold not by a single point but by a certain "height of metal".

I have a post last year when I found even some RCA socket for the 6C33C, although I thing it was made by Johnson and then rebranded

http://www.goodsoundclub.com/TreeItem.aspx?PostID=1093

Rgs,
The caT


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
06-11-2006 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 35
Post ID: 2510
Reply to: 1175
An good article about the 6C33C SET.

Here is an interesting article by Dmitry Nizhegorodov about his experiences to use 6C33C. (I do not know who Dmitry is) I disagree with Dmitry’s amplifier. Also, I very much disagree with his vision that an amplifier is juts a machinery where voltages and currents simply propagated according to the circuit’s rules. Of course all these things are important I did not see in the Dmitry’s article how it all realties to Sound.

http://www.geocities.com/dmitrynizh/ecc99-6c33c-se.htm

After all we do not built amplifiers for sake of amplifier but to serve the interest of Sound not the interest of the applied scientific theories. The theories of science just describe the events of Realty but they are not the Realty itself. In order an amplifier were able to operate by the language of Realty (instead of the language of Reality's description) the different design methods should be used. Anyhow, I feel that the Dmitry’s article and the Dmitry’s affords are very much worth attention as it electronically-mature written.

Rgs,
Romy the caT


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
10-02-2009 Post does not mapped to Knowledge Tree
Paragon
Posts 6
Joined on 09-13-2009

Post #: 36
Post ID: 11897
Reply to: 2228
Single-Ended 6336B Monoblock Project
fiogf49gjkf0d
Romy,any information out there for a set of Power / Audio output Trans/ for a Single-Ended 6336B Monoblock project?   Thanks, The Paragons
10-02-2009 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 37
Post ID: 11899
Reply to: 11897
No 6336B, just 6C33C. Nope, I did not deal with 6336B.
fiogf49gjkf0d
 Paragon wrote:
Romy,any information out there for a set of Power / Audio output Trans/ for a Single-Ended 6336B Monoblock project?   Thanks, The Paragons
The 6336B is twin triode, I have seen some peoples is it but I personally have no experience. The Milq versions that I tried personally was with 6C33C and with different DHT output tubes (but only for MF band).

The Cat


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
11-02-2010 Post does not mapped to Knowledge Tree
N-set
Gdansk, Poland
Posts 609
Joined on 01-07-2006

Post #: 38
Post ID: 14833
Reply to: 2215
6C33 at higher volatges & lower curents?
fiogf49gjkf0d
Risking that the topic has already been discussed and that I post
in a wrong category (Romy feel free to (re)move this post:
Is there any sense in trying 6C33 at higher volatges and lower currents?
More precisely, I've been wondering if  1/2 6C33 @ 300-320V, 30-40mA
is worth a try or the tube is bad non-linear there?
The amp I'm contemplating is a Romy-style-but-Push-Pull
amp for electrostatic headphones: something for the input - 6E5P - DC -  1/2 6C33 Push-Pull
stacked PSU's, the driver and the output PP choke loaded.
The load in a rough approximation is a 100pF cap (stator-stator).
Cheers,
jk



Cheers,
Jarek
STACORE
11-03-2010 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 39
Post ID: 14835
Reply to: 14833
The expressed perfect operation point is a fiction in a way.
fiogf49gjkf0d
 N-set wrote:
Risking that the topic has already been discussed and that I post
in a wrong category (Romy feel free to (re)move this post:
Is there any sense in trying 6C33 at higher volatges and lower currents?
More precisely, I've been wondering if  1/2 6C33 @ 300-320V, 30-40mA
is worth a try or the tube is bad non-linear there?
The amp I'm contemplating is a Romy-style-but-Push-Pull
amp for electrostatic headphones: something for the input - 6E5P - DC -  1/2 6C33 Push-Pull
stacked PSU's, the driver and the output PP choke loaded.
The load in a rough approximation is a 100pF cap (stator-stator).
 
N-set, I moved it into this thread; I hope you do not mind. I think the question is a bit irrelevant as there is absolutely no concept of linearity projected to operation point of a tube. Let me to explain.

You have an output tube, any abstract tube with any own characteristics. You have a load – a loudspeaker with all-imaginable parameters. Then you have single ended output transformer. Warn you right the way that I know nothing about PP and know nothing about OTLs. You pick transformer a ratio that would assure a proposer harmonic structure of your sound – in this care it will be proposer loading of your tube projected to given type of your load. Then you have a very simple task to assure that under the given load that your amplifies will give you max power. The concept of max power is an absolute key – there is no other considerations that need to be considered as at max power all seating come clean. You power your 6C33C with any voltage and any currant you wish but load it to your actual load and via your actual transformer. Then you begin to increase the input signal. You do it unit you have a first clip. The clip will be at the bottom of sinusoid or at the top of the sinusoid top of the sinusoid represent current, the bottom of sinusoid represent voltage (or opposite - I never remember). Let say the tope is current and the top got clipped. Then you increase current and reduce voltage. Playing like this you need to find a configuration when your tube has absolutely symmetrical clipping at tope and at bottom. In this configuration the tube will give up max power but the most important it will not run at voltage or at current starvation. Whatever voltage and current you will end up to be will be the “perfect operation point” for 6C33C”.

It might be a bit tricky and change of current will drive the plate impedance a bit and it would demand a change of transformer ratio, that would reset all your searches for symmetrical clipping. Any other tube, and particularly 6C33C will behave differently, particularly in beginning and in the end of it’s life. Still, the concept is there.

If the concept is there then why I declared the perfect operation point is a fiction. Because just naming the operation point meaningful only for standard typical load. What if I use my headphones from submarine sonar? They have impedance of 400R and to drive them with my custom transformer I would have symmetrical clipping at 300V and 25mA? Or alternatively: if I have 8 drivers of my woofer towers all connected in parallel and total impedance of 0.8R. I have a crazy transformer for this operation but the symmetrical clipping in this case mish take place at 50V and 1A of current on the plate. They all are hypothetic examples but all you need to understand is that in ANY situation of using ANY SET output stage it will be always the condition for symmetrical clipping and it will happen ONLY in one single dead locking between voltage and current at a given power rating of plate. THAT will be the ONLY ONE non-fictional operation point and this operation point worse only to you and your conditions.

Rgs, Romy the Cat
 


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
11-03-2010 Post does not mapped to Knowledge Tree
N-set
Gdansk, Poland
Posts 609
Joined on 01-07-2006

Post #: 40
Post ID: 14839
Reply to: 14835
Irrelevancy of my question and 6336
fiogf49gjkf0d
Thank you Romy, that was very informative and instructive!
I don't completely agree that OP is absolutely irrelevant for linearity.
It can happen that some regions, under given design constraints, will
not give "a proposer harmonic structure of your sound", irrespectively
of the load, simple because the characteristics bend oddly there.
Using your procedure one can (hope to) navigate away from them
but I have the HV constraint: without enough voltage swing and fast slew rate
electrostats sound anemic.
So my question was rather: do there potentially exist conditions
in the HV/low current region for getting a good harmonic structure,
under some optimal load?
However I've just realized that my question is irrelevant for other reason:
what works well for a SET, OPT coupled to a speaker is probably irrelevant
for a PP directly coupled to a GIVEN capacitive load...I have no possiblity to
manouver here with the load seen by the tube, e.g. by turns ratio.
The only parameter somehow tuning the coupling
is the plate load choke. For example a very quick (and perhaps very
flawed) Spice model showed that I would need around 30H choke per plate for a 6C33C
to have a +/- acceptable harmonic spectrum at 20Hz (OP is 320V/-150V/60mA, full tube, 100pF load).
Probably this is what one needs to make the load ellipse, posed by 100pF@20Hz
not to cut the tube at LF.

On a related tube: the mentioned 6336 looks very interesting, esp. that I could
use both halves for the PP. Any input on this tube highly appreciated (incl.
a good spice model for one section). Or better yet 6528..

Cheers,
N-set






Cheers,
Jarek
STACORE
11-03-2010 Post does not mapped to Knowledge Tree
Paul S
San Diego, California, USA
Posts 2,570
Joined on 10-12-2006

Post #: 41
Post ID: 14840
Reply to: 14839
Reverse Engineering
fiogf49gjkf0d
N-Set, I don't know about the 6C33C for tying directly to a load like that, if only because of its propensity to drift all over the place.  My own limited experience with this tube suggests that there is indeed a "sweet spot", but I am pretty sure you could not hold to it under the circumstances you propose.

It might in any case do to get the best "active" model you can for the speaker/load and then work backwards from there.  30H on a plate sounds downright bizarre.  How can this be a good idea?  Maybe better to stack more tubes?  6L6 or EL34?  Anyway, try to think like a capacitor...

Best regards,
Paul S
11-03-2010 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 42
Post ID: 14841
Reply to: 14839
Since I do not get electrostats...
fiogf49gjkf0d
N-set,

if you use PP directly coupled to a GIVEN capacitive load then you might find it interesting to talk with Dima. He told me about his very cool idea how electrostats (that are notoriously capacitive) might be driven directly from plate. The idea is not new but he proposed to make it differently and I did not see it done anywhere in a way he described. I did find it very elegant but since I do not like electrostats I did not make any further thinking or experimenting with it.

The Cat


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
11-03-2010 Post does not mapped to Knowledge Tree
N-set
Gdansk, Poland
Posts 609
Joined on 01-07-2006

Post #: 43
Post ID: 14842
Reply to: 14840
30H & EL34
fiogf49gjkf0d
Paul, would you think 6336 will be more stable/predictable?
Have you had any experience with it?

30H: sorry, forgot to mention that the PP pair is autobiased
with a common 900R resistor...

EL34: indeed it's a very popular tube for driving Stax electrostats (Stax T2, blue hawaii, no idea
on their sound though), but 5x or so lower the internal resistance of 6336, 6c33c etc is appealing, esp
with a fast driver. I would not like to go into 300B or 845 crazyness and 801A has a highish internal resistance.

Romy: sure, would love to talk to him! Due to the space I'm limited to headphones if I want
to listen to anything, and then electrostats seem to be the most promissing...a sort of a
negative deduction choice.

Cheers,
n_set



Cheers,
Jarek
STACORE
11-03-2010 Post does not mapped to Knowledge Tree
Paul S
San Diego, California, USA
Posts 2,570
Joined on 10-12-2006

Post #: 44
Post ID: 14843
Reply to: 14842
Headphones...
fiogf49gjkf0d
I don't know the 6336, but EL34 used correctly will sound like its driver tube, not to mention the comparative ease of PP scheming.  Sure, you need the voltage; but this to me only underscores the low ESR angle.  And who needs an inductor/transformer the size of a toaster?

Best regards,
Paul S
11-03-2010 Post does not mapped to Knowledge Tree
N-set
Gdansk, Poland
Posts 609
Joined on 01-07-2006

Post #: 45
Post ID: 14844
Reply to: 14843
6E5P stability
fiogf49gjkf0d
BTW, how stable 6E5P is? I'm asking as I was contemplating biasing
the output by the voltage drop on 6E5P's plate choke (DCR is say around 1k),
with only some minimal autobias on the output. The drawback is that any drift in the driver
will rebias the output (I'm all the time assuming DC coupling with stacked PS).

Cheers,
n-set 
 



Cheers,
Jarek
STACORE
11-03-2010 Post does not mapped to Knowledge Tree
deemon
Posts 24
Joined on 05-25-2004

Post #: 46
Post ID: 14845
Reply to: 14842
OTL electrostat amplifier
fiogf49gjkf0d
Yes , I told Romy about this funny idea , which is quite simple of course - to make a simple SRPP stage on special hi-voltage tubes , powered with high voltage supply ( 10000 V , for example ) and drive the electrostat panel directly from the output . The main thing here are the tubes , and I know only one Russian tube of this kind  - GP-5 , high-voltage regulator triode http://www.qrz.ru/reference/tubes2/type1/gp5.shtml , it has been used in old colour TV-sets , in the high-voltage regulator circuit for the picture tube . Of course , it must have many analogs in other countries ( and they can be even more suitable for audio amp ) , I don't know their names but I think that it's not a problem to find them . The main problem with this idea it that I haven't any electrostat speaker here to try , so it can be only a "theory" , of course ... And another point - we can use this ultra-high voltage amplifier only for big electrostats , like Quads , for example .... but for headphone electrostat we don't need such a voltage , so we can use the SRPP stage with more usual kind of tubes . 
Best regards 
Dima
11-04-2010 Post does not mapped to Knowledge Tree
manisandher
London
Posts 158
Joined on 09-05-2008

Post #: 47
Post ID: 14846
Reply to: 14845
EAR direct-coupled amp
fiogf49gjkf0d
Hi Dima,

Do you know if this is similar to the approach Tim de Paravicini took in his direct-coupled amp placed inside a Quad? I'm currently using Quad speakers (I gave up on the horn approach through sheer frustration) and it seems to me that this is a great way to go...

Mani.
11-04-2010 Post does not mapped to Knowledge Tree
Romy the Cat


Boston, MA
Posts 10,049
Joined on 05-28-2004

Post #: 48
Post ID: 14847
Reply to: 14844
About the 6E5P stability
fiogf49gjkf0d

 N-set wrote:
BTW, how stable 6E5P is? I'm asking as I was contemplating biasing
the output by the voltage drop on 6E5P's plate choke (DCR is say around 1k),
with only some minimal autobias on the output. The drawback is that any drift in the driver
will rebias the output (I'm all the time assuming DC coupling with stacked PS).
n-set,

the question about the 6E5P/6E6P stability I asked myself as well, what I was constructing with DHT version for MF. If you look MF channel of 6ch-Milq

http://www.RomyTheCat.com/TreeItem.aspx?PostID=10915

… then you will see that DHT tube is direct compiled and failure of driver tube will burn out  the output tube, in my case the rare and expensive 4V DH tubes. Yes, I put 100mA fast blow fuse on B+, just in case, but so far I did not have it blowing because a catastrophic failure of  driver tube. As you might see I use 6E5P/6E6P everywhere and since 2004 I went over I think 2-3 dozens of them. Some of them were very severely abused by a Moron with soldering gun (means me), I made many mistakes while I was experiencing with Melquiades and there were events when  6E5P/6E6P were running with no bias and as the result had anode with the temperature nears of Sun surface. Surprisingly, after cooling the tubes still forking fine, even though I did not put them in amps after that.

Uselessly what I take a new 6E5P/6E6P I test them.

I uselessly go the second posts of this thread

http://www.romythecat.com/Forums/ShowPost.aspx?postID=4805

…where the test data is located.

I usually test them at 4V bias, not the 3.4V as I wrote in the post. The 4V is how they run in Melquiades.  The good new tubes have GM around 19000, the well works tubes, after one year, have around 17000. The tube that approach 16000 are candidates for replacement but I sometimes replace tubes with 17000 GM. I look if the tube is old and if it has too much darkness built up on the glass. It might be not right way to do but I have many of those tubes – so why do I need to collect them? Live of high-end audio person is about who accumulate more tube before then die – I do not want to be a winner in this competition. I will be very comfortable if at the time of my final hart attack I will not have any tubes left….

Now, about the rejection rate. For all time I had one or two 6E5P/6E6P that I trashed right from beginning as they were not useable.  Another one or two tube did not give sufficient current. I generally would like to have 35Ma at 4V bias, if tune do not give it then I use it in A, B, C, D channels of 6ch-Milq. I need a full 35mA for single-stage amps of Melquiades. The 6E5P/6E6P are Russian tubes and as any Russian tubes they have notoriously wide parameters. The 6E6P-DR shall be military tube and shall have much less spin off the parameters but it is not what I see. What is good about those tube is that if you get one that operate as you want then it will most like to maintain this operation mode very stable. At least I did not have any single 6E5P/6E6P that failed on me after it was accepted on duty.

It might be no a lot but this is all data that I have. The tube is not used widely for a long time. The history of it’s use in high-end audio started in 1999 what Dima “discovered” it for audio. There is no known to me motioning of this tube in audio before.  If you plan to use 6E5P/6E6P then a tube tester (I am sure you have)  and it will be sufficient for initial filtering out the bad tubes, then they shall be working fine.

The Cat


"I wish I could score everything for horns." - Richard Wagner. "Our writing equipment takes part in the forming of our thoughts." - Friedrich Nietzsche
11-05-2010 Post does not mapped to Knowledge Tree
deemon
Posts 24
Joined on 05-25-2004

Post #: 49
Post ID: 14866
Reply to: 14846
Direct-coupled amp
fiogf49gjkf0d
 manisandher wrote:
Hi Dima,

Do you know if this is similar to the approach Tim de Paravicini took in his direct-coupled amp placed inside a Quad? I'm currently using Quad speakers (I gave up on the horn approach through sheer frustration) and it seems to me that this is a great way to go...

Mani.
Hi Mani , I don't know about it and never heard it before ... Maybe Romy knows more ? 
Best regards 
Dima
11-06-2010 Post does not mapped to Knowledge Tree
N-set
Gdansk, Poland
Posts 609
Joined on 01-07-2006

Post #: 50
Post ID: 14868
Reply to: 14847
Drifting and loading
fiogf49gjkf0d
Romy, I'm of course familar with your MF design and Jim's 6E5P curves is what
I base my considerations on at this moment (along with the corresponding Spice model).
Your DHT MF is quite self-balancing due to the large cathode resistor. What
I have in mind is a deifferent topology (derived from Chimera's Axiom
http://www.chimeralabs.com/images/axiom.sch.jpg), which looks like much more sensitive
to the drifts. I post my lamentable approach here (feel free to (re)move it)

(I have problems with uploading the gif to the gallery so I attach it).

Most of the output's biasing is due to a voltage drop
on a big, 1.2k,  DCR of 6E5P's chokes. Then follows a small selfbias with 150R,
which will not be efficient in correcting driver's drifts.
But what you say is encouraging:

"What is good about those tube is that if you get one that operate as you want then it will most
like to maintain this operation mode very stable."

As for the loading: thinking as a capacitor, voltage is what drives me, so I need
a constant current source. Paul, I'd have less problems accepting a huge choke
than e.g. a transistor CCS. I do in fact have in mind a choke on a huge amorphous core,
but heavily sectioned (and interlayed), so in reality a collection of small coils on a common core.
As for active tube loading in SRPP, I have no experience with
it, but IIRC in a more standard application of an input stage, it had some sonic problems,
which some people tried to link to a specific harmonic spectrum of SRPP. No idea if that
would apply to the current context though.

Best,
N-set






Cheers,
Jarek
STACORE
Page 2 of 3 (66 items) Select Pages:  « 1 2 3 »
   Target    Threads for related reading   Most recent post in related threads   Forum  Replies   Views   Started 
  »  New  6C33C myths: audio Moronometr...  Overdrive warning light...  Melquiades Amplifier  Forum     3  56367  06-22-2005
  »  New  The short "6C33C Survival Guide"...  Ac filament.....  Melquiades Amplifier  Forum     20  363209  12-18-2007
  »  New  A different breed of 6C33C amplifier...  6C33C mono block @ clipping...  Melquiades Amplifier  Forum     84  869110  09-06-2011
  »  New  6C33C socket types, where to get the ideal one?..  Not so much the material that matters to me but the pin...  Melquiades Amplifier  Forum     2  29463  01-09-2012
Home Page  |  Last 24Hours  | Search  |  SiteMap  | Questions or Problems | Copyright Note
The content of all messages within the Forums Copyright © by authors of the posts