My first design. PSE 211 to review. - Page 3 - diyAudio
Go Back   Home > Forums > Amplifiers > Tubes / Valves

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 7th April 2005, 11:09 PM   #21
EC8010 is offline EC8010  United Kingdom
diyAudio Moderator Emeritus
 
EC8010's Avatar
 
Join Date: Jan 2003
Location: Near London. UK
Regulator design becomes more critical for single-ended amplifiers. Push-pull and differential stages can tolerate "less than ideal" regulators better. Multiple LC stages are probably better for a SE amplifier where absolute voltage isn't critical.
__________________
The loudspeaker: The only commercial Hi-Fi item where a disproportionate part of the budget isn't spent on the box. And the one where it would make a difference...
  Reply With Quote
Old 8th April 2005, 01:34 AM   #22
MelB is offline MelB  Canada
diyAudio Member
 
Join Date: Feb 2005
Location: BC Canada
I'd go more capacitance less choke. I own one of those 193Q's. They are 21 pounds each. (Ouch!) I'm listening to my parrallel 300B amps now. There is no audible humm. 5AR4 => 20uF => 5H 200mA choke => 680uF => sound parts. Regulated filiment supply! MelB
  Reply With Quote
Old 8th April 2005, 05:25 AM   #23
PRR is offline PRR  United States
diyAudio Member
 
PRR's Avatar
 
Join Date: Jun 2003
Location: USA
> 100 Ohm change in load will cause abrupt changes in the distortion figures.

SPICE is sometimes pretty close, and sometimes it goes way off-track. NEVER believe it without question. We know a real tube will not show major change of performance for minor change of condition. Your posted numbers look "possible", so the way-out numbers are probably some type of mathematical collision. But even so, I would not put much weight on 0.05% versus 0.2%: both calculations may be wrong, even assuming the real tube matches the model (it never does).

You have, however, fallen into one SPICE-trap. In SPICE, parts are free. It is too easy to throw lots of big parts on the plan and get great numbers. As others have hinted, a few of those squiggles are BIG HEAVY COSTLY parts. That may be OK if you have the cash and space and floor-support. But you should also ask if you could apply that same money differently and get a better result.

Bad Example: the gross spec, >10W <1% THD, could be met with push-pull 6V6, overall feedback, much smaller audio iron, much-much smaller power transformer and filtering. Yes, I know you are not looking for yet another mini-Dynaco or small-Fisher, you want your wide-open SE triodes.

But here is another idea: today, caps are cheaper than chokes, and a big-C filter is usually your best buy. However caps are "cheap" only up to 450V, after that you must de-rate and series-string or buy huge oil caps. Also tubes up to 500+V can use efficient oxide-coated cathodes, over 1,000V secondary emission beats the oxide to death and we must use less efficient solid cathodes. Emission current is limited by low heater-cathode efficiency, so the way to big power is high voltage. Thorium in tungsten cathodes favor running the plate up in the 1,500-2,500V range. (The 211's plate rating, 1,250V, seems low, especially considering that at zero-bias it is almost dead up to several hundred volts.)

Also: I gather you never built a tube amp this big. 1,000V is NASTY DANGEROUS stuff. I took 600V through a finger once and it is still numb, 30 years later. 250V is also potentially fatal, but 600V or 1,000V really increases the odds of serious bodily harm.

I really think you should start at less-insane voltages.

10W SET.... well, four 2A3 or a couple.three 300Bs will do that at "only" 350V-400V supply. You self-bias it, not so much for stability but because the low-Mu 2A3 needs big grid drive (~60V) and your driver will need about 5 times the grid-bias to do the job well. Now your driver and output can eat the same power rail, eliminating quite a lot of parts. Your really low impedance 417 can easily drive four self-bias 2A3 grids and their gridleak resistor. The low load impedance, a few Kohms, makes output transformer winding very simple (winding capacitance hardly matters).

One objection is that 2A3 and 300 went from well-used to old-junk and bounced: are now cult-status, with high prices. (I remember when you could get baskets of 2A3 for $0.33 each, now $100-$500+).

Some bottleheads will sneer, but triode-strapped 6550 makes a respectable output device. No, it isn't a true triode, though I'm not sure the electrons can tell the difference. (Triode-strapped 807 was good enough for Williamson.) No, it isn't directly heated, a concept that went very far out of fashion before it became the "IN" thing again. A couple 6550 triodes will make a heck of a sound in any efficient speaker. Maybe not >10 Watts, but not many dB less. (BTW, I have not seen a SPICE model for a pentode, even triode-strapped, that seemed as accurate as eyeballing charts.) Being a cheapskate, I would think about 6SL7 to drive it: plenty of voltage-gain for CD source and 250K grid-loads.

I am no longer the heavy-feedback fan I used to be, but one thing feedback IS good for is turning low hum into very low hum. The compact affordable hi-fis of the 1950s would not have been possible without NFB. (But in retrospect, maybe they were TOO compact and affordable....)
  Reply With Quote
Old 8th April 2005, 06:52 AM   #24
diyAudio Member
 
audiousername's Avatar
 
Join Date: Jul 2004
Location: Melbourne, Victoria
Default Re: Elevated 417 filament

PSE 211 with a 1kV+ PS with four big, heavy filter chokes for a measly 16 watts or so is a little extreme. You might be better off with a cheaper, lower valve like a GM70 or something if you must have a transmitter valve SE amp. The lower valves allow you to swing closer to 0Va before drawing g1 current at the expense of more severe driving requirements. Even better, if you can find a high-ish and high-ish gm valve which tickles your fancy, you can have both!

Quote:
Originally posted by Radames
Tim, could you explain me how elevating the filament supply will reduce hum?
The idea is that there is a parasitic thermionic diode between the heater and cathode which is turned "on" while the heater is negative with respect to the cathode (after all, the heater is just a bit of hot wire). Elevating the heater to a higher potential is done to reverse bias this parasitic diode and hopefully reduce any hum which is induced by it.
__________________
Jason
  Reply With Quote
Old 8th April 2005, 01:39 PM   #25
diyAudio Member
 
Sch3mat1c's Avatar
 
Join Date: Jan 2003
Location: Milwaukee, WI
Send a message via ICQ to Sch3mat1c Send a message via AIM to Sch3mat1c
Default Re: PSU hum

Quote:
Originally posted by Radames
Hey Tim thanks for the reply.

I calculated that every LC stage of the PSUs will give me around 58dB of hum attenuation. So together would give more than 100, and that's what I wanted to get.

Using a single LC could let the hum be heard through the speakers if my calculations are correct. But may be that's all math rubble.
Lemme see here...

Oh, don't forget that hum in the driver will cancel hum in the output, just how much depends on relative loading, filtration, gain and so forth.

So anyways, 750VAC rectified is naturally 750Vrms, but most of that is 675VDC DC output (choke input filter, since C10 isn't enough except at low current). I don't know what ripple RMS looks like. Let's arbitrarily say it's 400V ripple at the rectifier. This hits the first choke and cap as a voltage divider, 10H to 51uF, reactance (120Hz) 7540 ohms choke, 26 ohms cap, for a ratio (I/O) of 0.0034, or 1.38V. Neglecting the second filter stage, this would appear as 0.78 *1.38 = 1.08V on the transformer. At 8 ohms output (you left it unmarked, so I'm making an assumption), this is 36.5mV ripple at the speaker. Full power of 15W or so into 8 ohms is 11V, for a noise floor of 304 times under full power output, or -24.8dB (-49.7dBV).

Not too great, could be worse, could be better. As mentioned, some amount of cancellation will occur (this can be calculated).

But choke input is insufficient voltage, something you seem to have overlooked; so pulling the first choke, leaving 51uF at the rectifier, 10H in series, and the other 51uF, will net probably 1.4 x Vrms = 1050V or so (may be lower depending on current draw; 51uF isn't much, I'd go for 100uF). Ripple is maybe 8Vrms, and well since it's the same filter as above, you can see already that hum is 8/400ths lower, sounds SFA to me!

Tim
__________________
Seven Transistor Labs
Projects and Resources
  Reply With Quote
Old 8th April 2005, 01:42 PM   #26
diyAudio Member
 
Sch3mat1c's Avatar
 
Join Date: Jan 2003
Location: Milwaukee, WI
Send a message via ICQ to Sch3mat1c Send a message via AIM to Sch3mat1c
Oh, and I was going to comment on a regulator:

Okay, so it can kill hum. So can't a simple linear power supply. It may be lighter, but not much lighter than a conservatively designed supply (audiophoolery does not apply to my posts). And, of course.... you have to ask yourself... are you feelin lucky? (No wait)... you have to ask yourself, do you want to waste all that voltage? An LC filtered supply wastes *no* DC voltage.

Tim
__________________
Seven Transistor Labs
Projects and Resources
  Reply With Quote
Old 8th April 2005, 02:37 PM   #27
Radames is offline Radames  United States
diyAudio Member
 
Join Date: Apr 2005
Location: New York
Default Regulator option/ PSU issues

Hey! This is awesome! Thank you all!

Ok, Sy convinced me yesterday to look up the 317 options and there I went with my HP calculator. After trying to understand Morgan Jones' regulators for a while (he makes it quite hard), I figured that I still need one LC and then the regulator. I would save 1 LC combo and >40lb!
The price of the whole thing, however, rises. As PRR well mentioned caps that can take more than 450V get expensive and the regulator option implies adding 2 more caps than in addition to the ones I have already. Weight is still an important improement yes.

The choice of the caps an 193Qs was governed by whatever I could fetch from the Partsconnexion site with enough rating to handle the voltage and current (234mA at the operating point and 1006VDC). The 193Q is rated 500mA 1000VDC, the lower model is rated 300mA but only 800VDC. But may be this does not matter much because there is no potential difference (or very little; set by the resistance of the choke) between the choke terminals... If that is correct I could use the 193P at half the weight and 15 bucks less.

The choice of caps was again to find stuff that would take more than 1006V across it. The only one I could find (at PC) that could take >1006V was that 51uF Solen.

I admitt, the PSU has many problems, mainly weight and cost.

Cheers!

Rada
  Reply With Quote
Old 8th April 2005, 02:41 PM   #28
diyAudio Member
 
audiousername's Avatar
 
Join Date: Jul 2004
Location: Melbourne, Victoria
The voltage rating of the choke refers to the rating of the insulation between the windings and core. You can place the chokes in the ground side of the filter to get around this problem.
__________________
Jason
  Reply With Quote
Old 8th April 2005, 02:56 PM   #29
Radames is offline Radames  United States
diyAudio Member
 
Join Date: Apr 2005
Location: New York
Default P-P vs SE and valve choice and modelling

PRR very well pointed that a P-P is a better alternative. Morgan Jones could not agree more and he convinced me in his book.

I am a true beginner and I chose SE due to the simplicity of the circuit topology and ease of the calculations. I actually could not quite figure the inner workings of the P-P topology. I admit I need to study it more and re-read Morgan Jones' book.

PRR, I don't use any SPICE (only in my food). I actually don't know how to use any SPICE software. I used SwitcherCAD just to draw the circuit because it seemed faster than doing it in Photoshop.

All my calculations and loadline choosing is done partly in an old fashioned way. Drawing a million loadlines, calculating all the tube parameters and the values of the elements on the circuit for each line. I choose the line looking at the THD figures, power and coupling elements to ensure good bandwidth.
Of course I don't do it with a pencil, I have a Matlab program that does all the calculations for me (I am good with Matlab, bad with pencils). The tube models are based on Norman Koren's equation. So far I have implemented only the triode model, and that's why there is no pentode in my design . I know, it is a stupid reason...

If anyone uses Matlab and is interested in having a look at my program, just e-mail me.

I looked at different tubes: 845, 2A3, 300B, 211, 45, 801A, 6C33C-B and 27 as output tubes. Due to either price or power output I ended only with the 845 and 211. The 211 won because I found I needed less voltage to get the same power, and the linearity was superior to that of the 6C33-CB. The 2A3 and 300B are very expensive.
For input tubes, the only ones I could find that allowed to maintain the 2 stage topology were the 417A and the 6C45-PE. The linearity of latter is rather poor though.

Cheers,
  Reply With Quote
Old 8th April 2005, 03:03 PM   #30
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
PRR: Sane advice. That's really the best way to go for 99.99% of the builders.

Radames, did you read through Maida's app note? It's pretty straightforward. The main thing is to not be afraid to use lots of protective diodes and understand that silicon things pop when you fiddle around with them carelessly. I've got a prototype running here at 700V using an NTE165 as the pass device; it worked fine with a MOSFET there, too, a 2N6770. I admit that I blew up a few devices during troubleshooting...
__________________
You might be screaming "No, no, no" and all they hear is "Who wants cake?" Let me tell you something: They all do. They all want cake.- Wilford Brimley
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Please review my design kilowattski Analogue Source 69 29th March 2011 05:54 PM
Wilson Watt Puppy Clone Design - Design Review Please soundengine355 Multi-Way 8 15th August 2008 02:07 PM
Design review? Sawzall Digital Source 0 6th July 2003 05:47 AM
Hi/Review Design Sch3mat1c Tubes / Valves 37 29th January 2003 02:30 PM


New To Site? Need Help?

All times are GMT. The time now is 03:01 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