DIY Audio Projects Forum
 NEW  Matt presents bias and operation data for the 6V6 tube in SE operation - 6V6 Single-Ended (SE) Ultra Linear (UL) Bias Optimization.

DIY Audio Projects Forum

Welcome to the DIY Audio Projects Message Forum. Use these forums to discuss Hi-Fi audio and to share your DIY Audio Projects. Registration is free and required to post messages and view the file attachments. Registration will only take a minute and registered users do not see any advertisements. After you have completed the online registration process, check your email (including spam/junk folder) for the verification email to activate your account. New members are under moderation - so your posts will not be visible until approved by a moderator. See the Read Me 1st, Forum RULES and Forum FAQ to get started on the forum.

It is currently 28 May 2020, 17:33

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 2025 posts ]  Go to page Previous  1 ... 58, 59, 60, 61, 62, 63, 64 ... 203  Next
Author Message
PostPosted: 29 Sep 2011, 10:26 
Offline
Project Author
User avatar

Joined: 04 Jun 2008, 20:59
Posts: 4162
Location: Arizona, USA
Hi Everyone, I attached the schematic for the KT120 version. I'll leave it to you to figure out how to power it. Remember to keep the heater circuit isolated from the ground and bias it at about 70-80 vdc. The circuit is an evolution of the smaller amps and not a redesign. Notice that there are changes in the driver stage (tube type more B+, changed resistors, different NFB values) these are quite important as they represent the best choices. You can use a 5751,but the residual distortion will be an huge 0.05% higher 8-) . The output transformer or equal is important as well. Do not use the smaller 60 watt Edcor, it isn't rated for the dc current (continuous in this amp at 310ma). The LM317s are connected pin for pin and can share the same heat sink which must dissipate around 12 watts (don't skimp on it). The 317s share a single sense resistor (separate resistors didn't improve anything). I hope to do a full write up on the amps in the near future. The more adventurous of you can get started early. The two prototypes are being circulated for critical evaluation (to see if Oddwatt Audio wants to issue kits on them - almost certainly will be a go). How do they sound - like the big powerful amps they are. Virtually identical to the latest version of the KT88 mono blocks (BTW I am working up a set of instructions on how to update early amps to the latest version). The latest version of the KT88s share the changes in the driver stage, but not the NFB, output trannie, nor CCS changes. Just make the driver like the one in the attached schematic and you will have the latest changes. :thumbsup: The driver tube is not a 12AX7 as listed in text on the schematic.

Attachment:
Odd-Block-HP Sept 9, 2011.PNG


Good listening
Bruce


This post has a file attachment. Please login or register to access it. Only Registered Members may view attached files.


_________________
Some of my DIY Tube Amplifier Projects:


Top
 Profile  
 
PostPosted: 29 Sep 2011, 19:16 
Offline
User avatar

Joined: 14 Feb 2010, 13:13
Posts: 701
Location: Lisbon, Portugal
Hi Bruce!

I wonder if these changes may also apply for the 5751 + EL34 set (my current situation).

Great work you´ve been doing!

I would like to give this a try...

Cheers,

Miguel


Top
 Profile  
 
PostPosted: 29 Sep 2011, 20:17 
Offline
Project Author
User avatar

Joined: 04 Jun 2008, 20:59
Posts: 4162
Location: Arizona, USA
Hi Miguel, Yes they can work with the KT77/EL34 versions although the difference in performance is much less. They work best on the KT88s and KT120s. It seems that they are a bit more particular in how they are driven. To test the KT88 values I used my reference amps that are wired to take both KT88 size and KT77/El34 size tubes. The improvements had a greater effect on the KT88s. Roughly a 0.1% drop in distortion a low levels and nearly 0.5% as it reached full power. For the smaller tubes the difference at low levels was about 0.05% (hardly worth the trouble and at max about 0.15%. The smaller tubes tend to be cleaner to start with so not getting a big improvement is not really a surprise. Since the modifications are small, personally I would do them. Would they be obvious when listening, perhaps not depending on your source material, speakers and most importantly your ears. They do however show up on the test gear.

All this does point out something... in any of my projects both diy and kits, I intend to post improvements as they occur in the evolutionary process. I don't consider a project ever done. :) I figure that particularly the diy community is always interested in tweaks that improve the equipment. For the kit builders I am in the process of preparing detailed instructions on what can be upgraded and I expect that a "parts" kit will become available to update the existing amplifiers.

Good listening
Bruce

_________________
Some of my DIY Tube Amplifier Projects:


Top
 Profile  
 
PostPosted: 30 Sep 2011, 00:18 
Offline
User avatar

Joined: 27 Sep 2010, 00:52
Posts: 22
Location: California
Hi Bruce,

This is truly inspiring. Thank you!

My first question is still regarding how you bias the power stage. You specified a 35-40V bias voltage, and 310mA Ia per pair of KT120 or 155mA each. I only have KT120 curves for either pentode or triode connection, but not ultra linear. The pentode one shows the last curve of 25V before floor. I would imagine a 35-40V curve would be near or at current cutoff (can only swing in one direction). But the triode curve shows an operating point of 35 or 40V at a reasonable mid-point for swings in either direction (class A1). It's correspondent Ia~150mA and Ua~420V. I wonder if this was how you determined bias (using the triode curve)? If not, how did you actually do it?

My other question is: mu for 12SL7 or 5751 is 70. Assuming your gain stage can get A~50. Given your power stage needs a 80Vp-p input to drive to full output, does that mean the gain stage has a sensitivity of 80/50=1.6Vp-p or 1.13Vrms? This hardly gives enough headroom for modern CD players. Too low for older technologies. Is my calculation wrong?

_________________
Striving for Eo=Ei*A


Top
 Profile  
 
PostPosted: 30 Sep 2011, 01:08 
Offline
User avatar

Joined: 27 Sep 2010, 00:52
Posts: 22
Location: California
Hi Bruce,
It's me the "curious George" again.
I remember reading somewhere, large power tubes like high current driver tubes to drive them. Have you heard of it? If it's true, the 12SL7 are 5751 are by no means high current (<3mA and <1mA respectively). Then 12SN7, ECC82 may come to picture (each up to 20mA). Another tube I have been told very good to drive large tubes like 300B is ECC99, up to 60mA, or even EL844/EL84, also up to 60mA). Of course their mu around 20 is low (back to my previous question about the gain stage not having enough gain, now even less.) So I am wondering if you consider it justified to add a higher current driver stage? For better drivability and for more head room (hopefully reaching an input sensitivity in the order of 0.1Vp-p, say first stage A=50, second stage A=16. Thus 80/50/16=0.1).
Just curios about what you think?

_________________
Striving for Eo=Ei*A


Top
 Profile  
 
PostPosted: 30 Sep 2011, 09:55 
Offline
Project Author
User avatar

Joined: 04 Jun 2008, 20:59
Posts: 4162
Location: Arizona, USA
Hi, In a way true, that is if you need current to drive the tubes in what ever design it is. All my designs use only voltage drive for the output tubes and a very low amount of current are needed. Generally only enough to handle the load of the next stages grid resistor (often a 220K). My designs are all what is commonly know as A1 class (no grid current drive), the other class is A2 in which grid current is needed. As a so-so generalization it is munch more common in large low gain power triodes than in pentode and beam tetrode applications. Sticking to A1 mode eliminates one stage in my amplifiers and my thought is the less stuff in the signal path the better. It does however cause some limitations. Mostly that the tube used as a driver must have a lot of gain and be linear over a wide range of signals.

Good thinking on the question, I'm sure others have had similar thoughts and not asked. In the beginning many folk thought the oddwatts could not work at all or if they did it would be rather poorly. 8-) Ask around, they work and work quite well.

Good Listening
Bruce

EDIT I didn't see the earlier question. You are approximately correct on the gain calculations. The amps require about 1 to 1.5 volts of input for full output. The expectation is that they will be used with a line stage preamp. The exception is the Poddwatt, that only needs about 0.5 volts of drive and is a stand alone device. BTW my CD player (OPPO83SE) delivers over 2 volts output and my phono preamp almost as much. I use a passive (no gain) preamp and I have no problems driving the amplifiers. As you would expect, the drive needed also depends on the tube type used. The Oddblocks can use everything form a 6L6GC to KT88s in the standard configuration (you do need to reset the idle current for the two sizes). The drive then needed can go from as low as +/- 20 to about +/-35.

The bias is automatically "set" by the constant current source. I set the current at the value I want and it does the rest. The values shown are measured ones not calculated. That is why there is a range. Different brand tubes end up with different values. (a KT88 from one company is not likely to match the specs or performance of one from another company) For U/L power curves and dissipation I use typically 80-85% of the combined values for anode and screen unless a specific tube max is given. This allows for good tube life.

_________________
Some of my DIY Tube Amplifier Projects:


Top
 Profile  
 
PostPosted: 30 Sep 2011, 10:41 
Offline
User avatar

Joined: 27 Sep 2010, 00:52
Posts: 22
Location: California
Thanks Bruce for the education!

I still want to know how you determine the bias level for a given power tube. In this case which KT120 curve (pentode or triode, or anything else) did you use? Was my "guess" of finding the operating point (in my previous post) close or far away? How did you actually do it? Thanks!

_________________
Striving for Eo=Ei*A


Top
 Profile  
 
PostPosted: 30 Sep 2011, 11:41 
Offline
Project Author
User avatar

Joined: 04 Jun 2008, 20:59
Posts: 4162
Location: Arizona, USA
Hi, The nice thing about the oddwatt amps is that you don't need to determine the bias at all. The design can be applied to pentodes, U/L or triodes equally well. The CCS will adjust the cathode voltage to make the idle current what ever you set it for. All you need do to determine the operating point is select the dissipation you want and a load impedance to match it based on the B+. The design of the circuit takes care of the rest. In most other amps you would need to determine the required bias from the curves. In oddwatts it doesn't matter what it is. This is contrary to about 99% of existing amplifiers and requires a rethinking of actually what goes one in tubes. A key thing to remember is that "bias" is the voltage between the grid and cathode. It doesn't matter to the tube much how you develop it. Since the grids are effectively at ground, then the cathodes are raised above ground by the CCS. The voltage it develops is a direct function of the current flow programed by the sense resistor. Since the tubes will conduct fully if you have no bias, the CCS keeps raising the voltage level until the current flow through the tubes is the same as was determined by the sense resistor. This takes place constantly. In a class A push-pull amplifier when one tube conducts more the other conducts less. Balance is maintained as the CCS won't allow either additional or less current flow. A consequence of this is the actual voltage across the CCS will vary in relation to the signal. It is what creates the drive for the second tube (remember the grid to cathode bit). As the driven tube is turned on more by an input signal, the voltage on the CCS tends to follow in doing so it makes the grid to cathode voltage on the second tube effectively more negative and reduces its current flow. If there is a negative signal to the driven tube, then just the reverse occurs. It is a see-saw effect. The original concept is at least 60 years old, but since they didn't have high precision constant current sources it was abandoned. The LM317 solves that problem quite nicely. If you look at the circuit in the cathodes the only adjustment is for balancing the current between the tubes. Even "matched" tubes do not balance precisely. With the control it is easy to get them to within 1% at low signal levels which are the most critical for normal listening. The balancing is done by shifting a small (2-4 volts) voltage (developed as an current times resistance drop) from one cathode to the other with the 25 ohm pot. Hope this helps. These amplifiers do not function like most others and it tends to confuse a lot of individuals. Like I said earlier keep asking the questions as it can help others understand as well.

Good listening
Bruce

_________________
Some of my DIY Tube Amplifier Projects:


Top
 Profile  
 
PostPosted: 01 Oct 2011, 06:00 
Offline

Joined: 24 Oct 2010, 07:05
Posts: 305
Hello Bruce,
While I agree with your explanations in general, I believe some of your words need some refinements. The question (as I understand it) was about how you selected the working point and working line of the power stage? You mainly explained what mechanism do you use to set already selected working point.
I'll explain. All tube gear works according to certain laws. You need to determine:
1. Working point, consisting of Ua-c, Ug1-c, Ia, Ug2-c (in case of tetrode and pentode) which are interrelated by some curves.
2. Working line, intersecting the working point on the curves with its slope determined by the load impedance of the stage.
3. In case of PP layout - how the "halves" of the stage is fed by the single input signal.
4. Limitations of the used parts (mainly tubes of course).
The working point could be set in different ways. I'm fully agree that CCS is the most smart way of achieving the result, because it automatically solves the point 3 in above list and allows very smart solving by only one part (an resistor in the CCS) the point 1. Solving, but not chosing:
gofar99 wrote:
All you need do to determine the operating point is select the dissipation you want and a load impedance to match it based on the B+.
This paradigm explains only the point 4 from the list above. Using the CCS explains setting for Ia and Ug1-c. But what about Ua-c, Ug2-c and the load impedance? In the case of CCS the things are complicated by dynamic nature of the load impedance (as soon as the CCS is a part of the impedance).
Then I'm in doubt about the fully correctness of your answer:
gofar99 wrote:
In most other amps you would need to determine the required bias from the curves. In oddwatts it doesn't matter what it is.

It matters. You have set the correct Ua-c, Ug2-c and transformer Ra-a and forgot about them, but it is the values which are define the "operating point". The fact that you haven't used the curves (KT120 don't have them either - correct curves I mean - all the same) doesn't mean that it doesn't matter.
The above reasoning allows to think, that there may be a room for imrovements in specifications comparing the current achievements. Bruce, you are the unique person with very high intuition, which allows you to achieve the remarkable result with the circuit (in my case I'm still experimenting with the power stage). But you should agree that CCS complicates the "determination things" a lot more, than the ways of practically implement the bias in common amps. I should agree nevertheless that your amp is extraordinary good.
wills wrote:
... which KT120 curve (pentode or triode, or anything else) did you use? Was my "guess" of finding the operating point (in my previous post) close or far away?

I afford to answer the question in a limit knowledge I have till now working with the tube. The widely circulated Internet curves are far from the real curves (I checked them with my tubes and a modified Russian tube tester L3-1). The only problem - I can't draw UL curves. :( The UL curves are definitely differ from plain pentode or triode curves and lies somewhere between them (look at KT88 family of curves for an example). All that considerations allows me to assume that your findings are not accurate.


Top
 Profile  
 
PostPosted: 01 Oct 2011, 11:44 
Offline
Project Author
User avatar

Joined: 04 Jun 2008, 20:59
Posts: 4162
Location: Arizona, USA
Hi Poty, Indeed it was not a complete answer. I tried to simplify the process. I'm sure you have studied (as have I) several references on the processes involved. The last one (in a different thread here) tried to put me to sleep on several occasions. There are many factors that relate to the operating point selection. However there are a much smaller number of variables that can be used to give quite good approximations of how a circuit will operate. Yes there is an intuitive factor involved as you observed that published tube curves and actual performance differ greatly. Now how I determine an operating point involves a consideration of practicality and also availability of components that can be obtained for diyers. I try to avoid specialty parts when possible. So, first I will not select a B+ in excess of 500 volts. Parts costs take a big jump if you go over that. Second for power amps I determine what output power I want. Since my personal thoughts are that class A amplifiers are the best (no flack please) and that push-pull have certain advantages with respect to harmonic distortion and they are what I design most of the time. I have found equally that U/L mode also significantly improves the designs - at a modest expense of output power. Now after determining desired output power and thus virtually forcing a choice of tube type ( because of tube limitations - you can not get 25 watts from EL84s and it is wasteful to use KT88s to get 5 watts) the next step is to make a few calculations regarding the input current needed to get the output desired. Since transformer efficiencies are seldom given, I use 90% as a value. At this point it is necessary to see if the tube type selected can handle the current and voltage. Here you need to put a bit of trust into the manufacturer's specs. In addition to dissipation I do check what they think the bias in a class A amp will be for the highest power they rate the tube at. I subtract this from the planned B+ to get a more accurate figure for voltage across the tube. Next comes the load for the tubes. Below is a description of how to do this.

General Output Transformer Considerations

A transformer will govern the current swing based on its impedance

For a 10K transformer each ma of change will cause a 10 volt drop in the effective B+

With the B+ the limiting factor of 525 volts and a cathode voltage of 25 volts the maximum swing is 500 volts in class A mode. Thus the current swing is then

500/10 = 50 ma. With a 5K transformer the result is 500/5 = 100ma

For push pull each side sees half the load in class A so for 500 volts into a 10K load the current will be the same as twice the 5K value for one tube or 200ma.

The next step is to calculate the amount of power dissipated Volts X current and see if it is below the maximum dissipation for the tube(s)

It is probably better to start in the reverse order and figure out what dissipation you want (perhaps 80% of maximum) the figure the current at the voltage available. You still will need to check the idle current to see that it is below the amount calculated.

So for a dissipation of 39 watts at 470-35 volts the current would be 90 ma. (as in the OddBlock Spec). The load needed to make that swing would then be 435/.09= 4800

For push pull class A it would be twice that value or 9600. Either 8K or 10K would be suitable.


The above process assumes that the transformer is made to spec. From this point on the CCS will do the housekeeping to make it work. Some hidden factors are: can the transformer handle the dc current, and is it really 90% efficient.

Good Questions. :thumbsup: I tried not to write a book, but the topic is a bit difficult to explain easily. Like I said earlier the last book I read on this spent nearly 100 pages on this topic spread out in several chapters. :sleep:

Good listening
Bruce

_________________
Some of my DIY Tube Amplifier Projects:


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 2025 posts ]  Go to page Previous  1 ... 58, 59, 60, 61, 62, 63, 64 ... 203  Next

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 9 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
DIY Tube Projects :: DIY Tube Amp Kits :: DIY Speaker Projects :: DIY Solid State Projects :: DIY IC / Op-amp Projects :: DIY Phono Projects :: DIY Cable Projects :: Hi-Fi Audio Schematics
© diyAudioProjects.com - Terms of Service - Privacy Policy