| arnoldc |
happy holidays!
i've been fighting with distortion on my prototype 807 amp. here are the details:
807 parameters (wired for tetrode):
B+ = 370V
Bias = -22
Cathode resistor = 500R, bypassed with 100uF
Screen voltage = 300V
i've tried near to a dozen variations of the parallel 12AU7 driver using tube cad and almost all resulted in distortion which is audible with my test monitor audio B2 speakers.
the latest version of the 12AU7 which seems to have the lowest audible distortion is as follows:
220R cathode, unbypassed
-2.3V bias
8K2 Rp
300V B+
where else should i be looking at?
thanks in advance.
PS.
there is no feedback as i don't know how to calculate for one. |
|
|
| SY |
Let me get a few details straight: you've got fixed bias AND a cathode resistor on the output stage? What's the idle current?
What's the topology of your driver? What are the DC voltages at each of the pins? |
|
|
| arnoldc |
| quote: | Originally posted by SY
Let me get a few details straight: you've got fixed bias AND a cathode resistor on the output stage? What's the idle current? |
it is around 44mA for the 807 (calculated). should i lower the cathode resistor to 400R or so?
| quote: | Originally posted by SY
What's the topology of your driver? What are the DC voltages at each of the pins? |
12AU7 wired in parallel. cathode to ground is 2.3V, plate to ground is 100V. |
|
|
| SY |
Toss the calculation- you're trying to get an amp to work, not a calculation. What is the actual 807 idle current? And why do you have both fixed bias and cathode bias?
The 12AU7 is going to have a hard time swinging a lot of volts the way you're running it. You might want to try a bigger plate resistor (like 25-50K), then go up on the cathode resistor a bit to try to hit something like 150V on the plates. |
|
|
| arnoldc |
hi sy,
i only have cathode bias via the 500R resistor.
let me go back to the workbench as measure the current. |
|
|
| arnoldc |
| quote: | Originally posted by SY
Toss the calculation- you're trying to get an amp to work, not a calculation. |
as measured, 44.3mA
sorry about the -22V i initially posted which misled you to fixed bias. |
|
|
| SY |
| OK, that's a reasonable idle current. I'd disconnect the output stage, stick a scope on the plates of your driver tubes, and see if they can actually swing the 20-30 volts you need. My suspicion is that the really low value of plate resistor is causing you the problem. |
|
|
| arnoldc |
oh, i don't have a scope :(
fwiw, in one incarnation, i have a 24K plate load on the parallel 12AU7 243V B+, 1K cathode resistor and that gave me 123V plate to ground and 5V cathode to ground. i still got distortion. :(
thank you for your patience. |
|
|
| SY |
| Can you draw up a schematic of the whole thing, from input to output? That would be helpful in the diagnostics. |
|
|
| SY |
| BTW, if you want to try designing and building your own amp, you need a scope and a generator. They're not expensive and they are as important as a drill and screwdrivers. Your problem could be something as simple as oscillation, but without being able to look at the signals, you're working completely blind. |
|
|
| arnoldc |
hi sy, my host does not allow image linking, so please click URL below.
807 schematic
i'm borrowing a scope and signal generator from a friend. |
|
|
| fdegrove |
Hi,
| quote: | | HTTP1.1 STATUS 403 Remote Access to this object forbidden This file cannot be directly accessed from a remote site, but must be linked through the Brinkster Member's site. |
Arnold, can you save the file to your local PC and upload it here?
Cheers,;) |
|
|
| SY |
I can't get that link to work- I get an error message that I need to be linked from a member's site.
Glad you're going into battle with the proper armaments! Look for oscillation. Look at each stage in isolation and see where overload points are. |
|
|
| SY |
| I wonder if your problem might be at the screen of the 807...? You don't have any kind of regulation there. Just for an easy diagnostic, tie the screen to the plate via a small (say 1K) resistor and see how things work in triode mode. |
|
|
| 316a |
Hello ,
The 12AU7 anode load is too low , also there is no feedback so I'd expect the high output impedence will react heavily with your speakers . The 12AU7 may not allow enough gain , especially with an unbypassed cathode resistor : suggest an EC8010 , 6N1P or E80CC instead , wrapping feedback from the output transformer secondary to the input stage cathode
316a |
|
|
| Gabevee |
Hey ArnoldC,
Try hooking the input to the 807 via the interstage capacitor, bypassing the 12AU7. Of course, remove it from the 12AU7 plate connection first. You should get a good enough output to hear if the output stage is distorting or not. Tetrodes have a high enough gain to drive with a small voltage. At least you can, without a scope, listen to see if that is the bad stage.
Otherwise, I think that the 12AU7 stage is the culprit. You are paralleling them, so the values for resistances for optimum performance should be halved. So, to get a decent response from the 12AU7, use about 5 mA per triode, or ten for paralleling. If you need a bias of 4 volts (optimum for 125 V at the plate), that makes the cathode resistor 4V/10mA or 400 ohm (V/I=R. Ohm's Law). The plate resistor should then be 125V (about half of the 250, or 245) divided by 10mA, or 12K5. This should give you optimum performance for the 12AU7 in parallel (I used the plate curves from the tube manual to arrive at these calculated values) .
One other thing... use a capacitor on the input. There is perhaps a load down of the grid circuit being imposed by your source. This will change bias and create distortion. Also perhaps a 220K grid resistor instead of 100K.
My :2c:
Gabe |
|
|
| burnedfingers |
How about scrap the 12AU7 and use a 5687 in its place? You ought to be able to drive that puppy with a 5687.
That circuit kinda looks like the LAF SE807RH that can be seen at WWW.diyaudio.8m.com
J |
|
|
| NickC |
Arnold
From looking at your output stage I feel that your screen biasing looks imcomplete. There should be a capacitor between screen and cathode as i recall. But conventional you could utilize a voltage divider to bias the screen to. Whats your screen voltage? from the big 50k resistor, screen current should be really low this would not load the tetrode right. They need to be fed with suffice current with respect to plate voltage and grid voltage. You have to view the pdf for 807. My i suggest looking for a super big pdf around 6-8mb. All you need to know about 807. I think should be available at retrovox website. Great site about screen voltages and care for them and understanding them
http://www.webace.com.au/~electron/tubes/index.html
As mention the tetrode as high gain and high output impedance as well.
changed grid to cathode. Was not thinking when typing that but thats what i meant bummer:smash: |
|
|
| SY |
| Nick, that's the point I was trying to get at a few posts back. But any bypass (or, much better still, regulation) has to go between screen and cathode. If the cathode is well-bypassed, it can go between screen and ground. But it seems to me that it ought to be a low-Z source. |
|
|
| arnoldc |
| quote: | Originally posted by 316a
The 12AU7 may not allow enough gain , especially with an unbypassed cathode resistor : suggest an EC8010 , 6N1P or E80CC instead , wrapping feedback from the output transformer secondary to the input stage cathode
316a |
hi 316A, in one incarnation, i did have a cathode bypass using 100uF (1K cathode resistor). the plate load was 24K, B+ of 243V, plate to ground at 123V.
| quote: | Originally posted by Gabevee
Hey ArnoldC,
Try hooking the input to the 807 via the interstage capacitor, bypassing the 12AU7. Of course, remove it from the 12AU7 plate connection first. You should get a good enough output to hear if the output stage is distorting or not. Tetrodes have a high enough gain to drive with a small voltage. At least you can, without a scope, listen to see if that is the bad stage.
Otherwise, I think that the 12AU7 stage is the culprit. You are paralleling them, so the values for resistances for optimum performance should be halved. So, to get a decent response from the 12AU7, use about 5 mA per triode, or ten for paralleling. If you need a bias of 4 volts (optimum for 125 V at the plate), that makes the cathode resistor 4V/10mA or 400 ohm (V/I=R. Ohm's Law). The plate resistor should then be 125V (about half of the 250, or 245) divided by 10mA, or 12K5. This should give you optimum performance for the 12AU7 in parallel (I used the plate curves from the tube manual to arrive at these calculated values) .
One other thing... use a capacitor on the input. There is perhaps a load down of the grid circuit being imposed by your source. This will change bias and create distortion. Also perhaps a 220K grid resistor instead of 100K.
My :2c:
Gabe |
hi gabe, i can try that (direct to the 807) and i'll configure the driver as you said.
| quote: | Originally posted by burnedfingers
How about scrap the 12AU7 and use a 5687 in its place? You ought to be able to drive that puppy with a 5687.
That circuit kinda looks like the LAF SE807RH that can be seen at WWW.diyaudio.8m.com
J |
hi j, i was about to scrap the 12AU7 and put in a 417A. i didn't have a problem with the 417A when i used it direct-coupled to a UX245. it sounded clean at full volume right after i powered it up.
i checked the LAF 807 and the grid stopper 200R and grid resistor 470K are the same. geez, this amp should sound ok then huh? they have a plate-to-plate resistor of 100K and labelled Rfb. it's an unconventional feedback compared to most i've seen.
| quote: | Originally posted by NickC
Arnold
From looking at your output stage I feel that your screen biasing looks imcomplete. There should be a capacitor between screen and grid as i recall. But conventional you could utilize a voltage divider to bias the screen to. Whats your screen voltage? from the big 50k resistor, screen current should be really low this would not load the tetrode right. They need to be fed with suffice current with respect to plate voltage and grid voltage. You have to view the pdf for 807. My i suggest looking for a super big pdf around 6-8mb. All you need to know about 807. I think should be available at retrovox website. Great site about screen voltages and care for them and understanding them
http://www.webace.com.au/~electron/tubes/index.html
As mention the tetrode as high gain and high output impedance as well. |
hi nick, i checked the morgan jones book and saw a capacitor to ground on the screen as well. my screen voltage is 300V (on the limit). i tried a voltage divider that gave me 200V for the screen but i didn't like the sound. i'll try it again and dial in for 250V at 2.5mA (screen current demand). yes, i do have that monster PDF for the STC807.
thanks to all who have assisted. several things i need to do now.
a) get a scope and sig gen ;)
b) learn more about 12AU7 op points
c) fix the screen grid voltage supply
d) dump the 12AU7
e) dump the 807 and stick to triode :xeye: |
|
|
| arnoldc |
| quote: | Originally posted by Gabevee
Otherwise, I think that the 12AU7 stage is the culprit. You are paralleling them, so the values for resistances for optimum performance should be halved. So, to get a decent response from the 12AU7, use about 5 mA per triode, or ten for paralleling. If you need a bias of 4 volts (optimum for 125 V at the plate), that makes the cathode resistor 4V/10mA or 400 ohm (V/I=R. Ohm's Law). The plate resistor should then be 125V (about half of the 250, or 245) divided by 10mA, or 12K5. This should give you optimum performance for the 12AU7 in parallel (I used the plate curves from the tube manual to arrive at these calculated values) . |
hi gabe, i entered your op point into tube cad and this is what i got :) |
|
|
| jlsem |
| quote: | | e) dump the 807 and stick to triode |
Yes, a single-ended 807 in BPA mode into nearly any load is an intrinsically high-distortion output stage, Tubecad notwithstanding.
John |
|
|
| arnoldc |
| quote: | Originally posted by jlsem
Yes, a single-ended 807 in BPA mode into nearly any load is an intrinsically high-distortion output stage, Tubecad notwithstanding.
John |
john, would you advise using an EL34 in single-ended pentode instead? i was originally prototyping the EL34 when the looks of the 807 caught my fancy. |
|
|
| burnedfingers |
quote:
e) dump the 807 and stick to triode
Well, I'll give you my experience with a SE 807 I purchased off the internet. It was a custom "one of a kind" amplifier that was advertised at 10watts per channel in triode mode. When I first took delivery of it there was very little gain before distortion. I sent the unit back and received it shortly after and it had a lot of gain but still fell far short of the advertised output. It put out a big .6 of a watt with a 1K sine wave and driving a 8ohm load. To be honest it was .6 of very very sweet sound. I sent it back because I didn't feel that I had received my moneys worth. It has taken over 1 month and threats of legal action before my refund came to me.
Dump the 807?
Well, personally if I felt that I wouldn't be collecting parts for an 807 version built by me.
We are fighters here
I know that I don't have the tube knowledge that others have here and I don't/won't pretent to. I do however have belief that if united we can accomplish anything here. I would ask that a solution be found first thus furthering education and knowledge for everyone even remotely interested in your project before moving on and scrapping the 807.
Just my .02
J |
|
|
| arnoldc |
hi j, i think i had read your post about your 807 purchase. looking at the 807 manual, i don't find any indication that would allow me to achieve 10 wpc in TRIODE. that's why i'm trying to get this thing wired in TETRODE for the 10 wpc which is achievable according to specs.
yeah, we are all fighters here, but i'm not armed yet :xeye: i haven't got the scope from my friend :bawling:
will all those who have ventured in 807 please come forward with your operating point for most power at lowest distortion. :) |
|
|
| Gabevee |
| quote: | Originally posted by arnoldc
hi gabe, i entered your op point into tube cad and this is what i got :) |
Hey ArnoldC,
I can barely read that picture. What is it trying to say?
Gabe |
|
|
| arnoldc |
hi frank,
that schematic is the same as the one posted by burnedfingers earlier. ironically, my chosen parts are almost the same as that one.
hi gabevee,
sorry about the small picture. attached, please find a better one. incidentally, i already plugged in into my amp and though it sounded better, i still got distortion in the output. so i was thinking i have a wrong operating point for the 807. a friend recommended that i wire the 807 in triode first, which i avoided because i wanted more power. |
|
|
| filippo22 |
IMHO the parallel 12AU7 isn't the best way to drive the 807... maybe a cathode follower or a totem pole. In my 807SE I used a PCL82 for driving the 807 into the positive grid region (A2) without big trouble.
ciao
Filippo
www.audiofanatic.it |
|
|
| fdegrove |
Hi,
Note also that Filippo used a separate screen supply which is how it is best done.
This is a snippet from a STC datasheet:
Cheers,;)
EDIT:The line at the bottom should read:
In cases where.......a potentiometer....and not by means of a series resistor. |
|
|
| Gabevee |
| quote: | Originally posted by arnoldc
sorry about the small picture. attached, please find a better one. incidentally, i already plugged in into my amp and though it sounded better, i still got distortion in the output. so i was thinking i have a wrong operating point for the 807. a friend recommended that i wire the 807 in triode first, which i avoided because i wanted more power. [/B] |
Thanks!
BTW, Someone else posted that you need to put a bypass on the grid 2 circuit. What happens is that without the bypass capacitor, the grid 2 sees a fluctuation like that at the plate. In other words, it acts like a negative feedback grossly reducing your output, because it interacts with the input (grid 1), since grid 2 can also be an input grid, hence the distortion. I also made the same error when playing with my 6L6 SE amp. I bypassed the grid 2 and boy what sound! If that doesn't help, then dare I say the output transformer may be bad.
But, glad my suggestion helped make it sound better!
Gabe |
|
|
| burnedfingers |
Gabe
The unit I had used a 100 ohm resistor from grid 2 to the plate for triode mode. Should this have been bypassed or am I misunderstanding you?
The unit I had didn't have the cathode resistor bypassed either. Since I only could get .6watt out of it do you have any idea how much would have been gained by bypassing it too?
J |
|
|
| Gabevee |
| quote: | Originally posted by burnedfingers
Gabe
The unit I had used a 100 ohm resistor from grid 2 to the plate for triode mode. Should this have been bypassed or am I misunderstanding you?
The unit I had didn't have the cathode resistor bypassed either. Since I only could get .6watt out of it do you have any idea how much would have been gained by bypassing it too?
J |
Hi Burnedfingers,
When I put the bypass, power and subsequently volume went up to 6 watts... from 1.5 watts. I never had any luck using my 6L6's in triode mode. Not much output for any real comparison. Well, compared to a real triode like the 45 or 300B, anyway
For a triode connected tetrode/pentode, you do not bypass the 100 ohm resistor. One wants to have the screen grid to pretty much add to the plate with regards current flow. Bypassing will merely make it a tetrode with a very high voltage on the screen grid.
Gabe |
|
|
| richwalters |
hey there........there's some interesting fodder in SOUND PRACTICES magazine, fall 1992 quote <single 807 amplifier>; A cure for triode fever; by Gordon Rankin, Wavelength Audio. I think it's triode electronics website. I cannot give direct links as my system goes crazy.
There seems to be alot of info, but in all cases when we call the 807 a pentode, the g2 needs a stabilised source. The rest of the posters say same psalm....
rich |
|
|
| jlsem |
| quote: | | This is a snippet from a STC datasheet |
Please note that into the 6K load, nearly all of that lovely 12% harmonic distortion is 3rd order. The lower impedence load of 2.5K gives mostly 2nd harmonics. No offense to anyone here, but I can't imagine a single ended multi-grid tube amp ever sounding any good at all, unless it used an ultra-linear output transformer. Which leads to my next suggestion: The Lundahl website features a link to an EL-34 UL SE amp that looks well thought out and may well give you the ten good sounding watts you are after for a reasonable price.
John |
|
|
| richwalters |
The 807 should give 10 watts (60mA anode I) output in true tetrode mode with 450V B+ o/p tranny and stabilised 250V on g2 with 6Kohms o/p tranny....I've done such an animal!
Looking at the 807 charts, g1 is then lim to approx 20V onset clipping.....that leaves nearest 390 ohms for cathode resistor........ Someone else must have cobbled and got the same ????
rich |
|
|
| fdegrove |
Hi,
| quote: | | Someone else must have cobbled and got the same ???? |
Yep...
The problems with the original circuit are manyfold in that the anode resistor for the 12AU7 are too low in value, the cathose resistor for the 807 seems off too and the screengrid is swinging along with the plate....Add to that that the 807 likes a higher B+ and you get the picture: distortion.
Cheers, ;) |
|
|
| richwalters |
Frank......for matter of curiosity, have you got the spec for that Aleksandar kitic RH807 amp ?
rich |
|
|
| fdegrove |
Hi Rich,
| quote: | | Frank......for matter of curiosity, have you got the spec for that Aleksandar kitic RH807 amp ? |
If it's not on the website then, no, I don't.
Having seen other designs by him I take it distortion figures are acceptable.
Cheers,;) |
|
|
| burnedfingers |
Rich
Would you mind posting the schematic for your 10watt beast?
J |
|
|
| burnedfingers |
Looking back thru the notes I made on the 807 I owned for a short while I found differences that I will list.
Where your amp has the 470K resistor from grid 1 to ground mine had a 680K. Yours has a 500ohm cathode resistor and mine had a 250 ohm and 64mA of bias. Mine had no bypass cap and yours does. Mine had no 200 ohm resistor to G1 either.
Interesting to note that when mine had a 500ohm bias resistor it was more distorted. Mine didn't have coupling caps either between stages.
Mine couldn't get past that magic watt either. Wait!!! Anyone have Joel's email address? He likes those amps less than 1 watt. They go with his Altec speakers.
J |
|
|
| burnedfingers |
Quote:
The problems with the original circuit are manyfold in that the anode resistor for the 12AU7 are too low in value, the cathose resistor for the 807 seems off too and the screengrid is swinging along with the plate....Add to that that the 807 likes a higher B+ and you get the picture: distortion.
I can hear the wheels turning in Frank's head and I believe the answer is near. Maybe even a full blown schematic headed our way.
J |
|
|
| fdegrove |
Hi,
| quote: | | I can hear the wheels turning in Frank's head and I believe the answer is near. Maybe even a full blown schematic headed our way. |
You give me way too much credit, Joe...
Sure, I see it in my head already but that's easy for me, I don't have to build it, neither do I need to fund it...
Thanks for the nice compliment anyway...
When the person starting the thread wants to build a very good SE amp he'd better realise that he can easily end up spending alot of money even if the hand holding is free.
If memory serves Thorsten Loesch says the exact same thing...
SE is not the free lunch some people think it to be....I don't mean the member with the same initials either.:D
Cheers, ;) |
|
|
| arnoldc |
based on the rich input i've seen so far, my amp was obviously ILL-designed.
i already had the G2 resistor bypassed to ground, and my findings were:
- it sounds awful with 500R cathode resistor,
- i substituted 390, then 270 and heard it sounded better with 270 cathode resistor.
my next steps are as follows:
- raise B+ to 450 volts,
- use regulated supply for G2,
- use 417A driver, KYW gave me a nice operating point for that tube which sounded excellent on my UX-245 direct-coupled amp,
- get myself an oscilloscope! the signal generator is already sitting here doing nothing :(
i was nowhere in tubeland when the sound practices and similar magazines were available. i despise the day when WE300B (1940's) were available in the philippines like apples and costs so cheap, and all i do is build (not design) solid state amps as a hobby. :xeye:
can i series VR tubes 0D3 (150V) and OC3 (105) to get 255V? and is 255V workable instead of 250V for the G2?
i can't thank all of you enough for the experience and knowledge shared on this thread.
happy new year to all!
ps.
joel... joel tunnah? he's in audio asylum... |
|
|
| Gabevee |
OK folks, I have a couple of comments, just one being a minor complaint.
Someone here said that tetrode or pentode SE amps sounded bad. and not worth building unless using UL connection. Holy mmoly.
The first two SE amps I ever heard were pentode and both sounded excellent, that with garden variety mass produced iron. The first was a 6BQ5/EL84 one, and the next was a 7868 one.
I get rave reviews for my 6L6 SE amp. Granted, it does sound a bit better UL, but it is a kicker even without UL. That with the Hammond OPT. The One Electron has no UL tap and is superior sounding. And I do have several other amps of various components and configurations to compare, including the venerable 300B.
So, ArnoldC, don't be too discouraged by your bad experiences now. You are learning from them. The end result will be vastly superior to any mid level solid state amp you have ever built and heard, and at the worst (once distortion clears up) equal to high end SS stuff.
Keep at it. You won't be disappointed.
Gabe |
|
|
| burnedfingers |
arnoldc
I will have to agree with Gabe. Your 3/4 there now so why think about scrapping it now? Before anyone gets carried away with the merits of the Solid State amps lets review several things. Those fantastic distortion figures that are posted are obtained mainly at or near full output. Those figures are very dependant upon the bias setting of the output stage. Performance changes with the age of the unit/parts inside.
What can we actually hear in terms of distortion? It is said/written that we don't really hear distortion until it reaches 1%. Before anyone gets carried away with distortion figures around 12% one should realize that the amp is probably not going to be used in that range of output and therefore won't experience those distortion figures.
I have owned/repaired/played with some of the most expensive Solid State amps on the market but I have never experienced the realism displayed by tube amps.
Just my .02, ducking for cover now.
J |
|
|
| arnoldc |
thanks for the encouragement guys ;)
when i get this to work, it will be posted here with measured figures, from input to output.
it's 2004 over here, happy new year! |
|
|
| burnedfingers |
arnoldc
I started my 807 SE this afternoon so maybe we can compare voltages and see what comes of it. I will cheat and use a variable supply so it will be easier to change the old voltages.
Mine is using a 6SL7 and 6SN7 for the front end and will use some HAmmond 125ESE's.
J |
|
|
| fdegrove |
Hi,
As it happens I have a copy of the Gordon Rankin article.
If there's enough interest I'm willing to post the details, probably even the circuit diagram.
Cheers and Happy New Year to everyone, ;) |
|
|
| burnedfingers |
Hi Frank
I for one would be most interested in the article and the schematic. I would be interested in your comments also.
Happy New Year!
J |
|
|
| fdegrove |
Hi,
| quote: | | I for one would be most interested in the article and the schematic. I would be interested in your comments also. |
Will do...When the bublles have cleared up...
Cheers,;) |
|
|
| arnoldc |
| quote: | Originally posted by burnedfingers
arnoldc
I started my 807 SE this afternoon so maybe we can compare voltages and see what comes of it. I will cheat and use a variable supply so it will be easier to change the old voltages.
Mine is using a 6SL7 and 6SN7 for the front end and will use some HAmmond 125ESE's.
J |
hi j, i was thinking to use my variac for the screen grid supply. however, firing up the B+ (i have standby switch) and the variac (also has standby) could be a hand coordination challenge. i was wondering what will happen to the 807 tube if one of the other was on first.
i also have it wired using 125ESE, though i have one pair of custom transformer (5K) and a pair of Tango U-808 at my disposal.
will you be using the 450 Vp, 250 Vg2, -30V bias, 60mA operating point as suggested by the others in this thread?
hi frank,
whenever you're ready :) |
|
|
| burnedfingers |
arnoldc
I'm going for triode mode first with 250 on the plate and G2 strapped to the plate via a 100ohm resistor. If i get it to run then I will take G2 to a regulated 250v B+ and try it.
On the Hammond 125ESE transformers the primary has a blue and a brown. Does the blue go to B+ ?
J |
|
|
| arnoldc |
| the blue goes to the top cap, and brown to b+ |
|
|
| burnedfingers |
Quote:
hi j, i was thinking to use my variac for the screen grid supply. however, firing up the B+ (i have standby switch) and the variac (also has standby) could be a hand coordination challenge. i was wondering what will happen to the 807 tube if one of the other was on first.
Your variac is running a DC source correct? I would set the screen voltage roughly at 250 volt with the variac. I don't think there is much current draw on the screens so that shouldn't be a problem. I'm not an expert but I don't thin the .5 difference will matter much in getting both power supplys turned on.
Will let you know shortly what I come up with. Thanks for letting me know the correct color for B+ as mine isn't marked.
J |
|
|
| fdegrove |
Hi,
| quote: | | can i series VR tubes 0D3 (150V) and OC3 (105) to get 255V? and is 255V workable instead of 250V for the G2? |
You certainly can and this is exactly what Gordon Rankin did for his 807 SE amp.
B+ is 450V for the plate supply in his amp so he uses a 10K/10W resistor to drop the voltage before it goes in the VR stack.
Cheers,;) |
|
|
| arnoldc |
| quote: | Originally posted by burnedfingers
Your variac is running a DC source correct? |
yes, it is.
| quote: | Originally posted by fdegrove
You certainly can and this is exactly what Gordon Rankin did for his 807 SE amp.
B+ is 450V for the plate supply in his amp so he uses a 10K/10W resistor to drop the voltage before it goes in the VR stack.
|
hi frank, thanks for the info. |
|
|
| fdegrove |
Hi,
| quote: | | hi frank, thanks for the info. |
You're most welcome...I'll try to help you anyway I can.....
I started to draw the regulator this afternoon, just to let you know I'm working on it.
All the best for 2004,;) |
|
|
| arnoldc |
frank that's :cool:
on a related note, is there a way to use a tube to regulate voltage other than VR tubes?
i think there was a circuit in this board that does just that. anybody remember that circuit? |
|
|
| arnoldc |
| i have one Philips PCL82 here that i'd like to use as regulator. can you point me to the right place? |
|
|
| richwalters |
Hi H.New Year to all......re Gordon Rankin's article;
the best I can do is this website:
Sound Practices Index
... Hiroyasu Kondo A Single 807 Amp -- Gordon Rankin ... Brook Amplifiers Meet the Tube: The Trusty 807 Understanding Output ... Joe Roberts Building SE Amp Kits -- Seth ...
https://secure32.softcomca.com/sout...udio_com/bac...
Apparently the issue 9 is out of print......I cannot go deeper with system crashing.
rich |
|
|
| burnedfingers |
Update
In triode mode I ended up getting a mighty 1 watt out of her. Thats .4 more watt than the one that I purchase then sent back.
More later today
J |
|
|
| fdegrove |
Hi,
| quote: | | i have one Philips PCL82 here that i'd like to use as regulator. can you point me to the right place? |
You could use a series regulator but I really don't see the need.
Here's the shunt reg as per Gordon rankin:
The 450VDC B+ is derived from a 350-CT-350 175mA xformer feeding into a 5AR4 rectifier, output goes to a 50µF/500V cap followed by a 5H 150mA choke and finally into another 50µF/500V cap.
Cheers,;) |
|
|
| arnoldc |
hi folks,
i think i had it sorted out...
i hooked it up to my 16-years old BOSE 301 Continental and the sound is so nice, no distortion even though i reckon this speaker is lowish in sensitivity at around 86dB. i thought i was hallucinating, i'm using a DACT attenuator and i cranked up the volume to maximum and i still don't hear the nasty distortion as before. so i hooked up the amp to a pair of Monitor Audio B2 speakers. still, the sound is clear.
it has a character of its own, very likeable. the midrange is surprisingly captivating and has a bottom end that is convincing even though i'm using just a Hammond 125ESE.
i don't have a way to measure output power though.
i believe there are more adjustments necessary to squeeze more from this tube as i'm operating on 44.5mA and i'd like to go up to 50mA.
tomorrow i will hook it up to my main speakers, a pair of Infinity Kappa 200.
i would like to thank all those who spent their time sharing their experience and knowledge in this thread.
i would like to acknowledge full credit to Gabe for the 12AU7 operating points, which were used in the prototype, pictured below. |
|
|
| arnoldc |
here is the updated schematic.
a friend who will be dropping off the 0C3 and 0D3 was not able to come by so i went ahead and got me some Zener diodes - 1 x 51V, 2 x 100V all 1W. i also grabbed some 0.1uF 630V Mylar capacitors and paralleled it with the Zener diodes, as can be seen in the schematic.
i am still wondering why my Screen voltage across the series of Zener diodes is at 261V though. could be the caps?
the voltage readings were the actual voltages that was measured tonight, taken about 45 minutes of listening to the amplifier. i was completely amazed at the balance of voltages between the two channels now. prior to the modifications, one channel was off by as much as 3V on the cathode voltage. now, one is 17.8 and the other is 17.6 807 cathode to ground.
any other suggestions about this amplifier is most welcome. |
|
|
| richwalters |
I'm not a high voltage zener fan....noisy, and considerable voltage drift with temperature. I tried zeners in my tube amp and soon abandoned them.....
Go by the stabiliser circuit..... Frank posted.... using OA'S etc......great blue glow and discussion piece. Remember, tempting may it be; don't shunt a stabiliser tube with a medium uF capacitor or it will oscillate.
rich |
|
|
| Gabevee |
| quote: | Originally posted by arnoldc
hi folks,
i would like to acknowledge full credit to Gabe for the 12AU7 operating points, which were used in the prototype, pictured below. |
Gee, you are making me blush!! Stop it!:blush: ;)
As for the 125ESE transformers, they do sound OK. The 300 B makes them go down to 40 Hz!
But when you get a pair of the Hammond 1627SE's look out! Those bad boys sound terrific!
See, you do not sound disappointed. I am sure you are delighted with all your good, hard work.
Happy listening!
Gabe |
|
|
| burnedfingers |
arnoldc
Very nice job! Your amp looks terrific. Glad to hear that you have her running. They do have a sound all their own I think. I fired up mine on the bench today and listened to some music. I have the grid supplied from one source and the plate voltage from another. Triode sounded like **** when I started pushing it.
I ended up using 2) 6SN7's for the front end and I still can't get more than about 3watts RMS into a 8ohm load with a 1K sine wave. I have plenty of drive voltage to the 807 so I am scratching my head on this.
I don't know how people get 12watts RMS out of one of these.
J |
|
|
| arnoldc |
| quote: | Originally posted by richwalters
I'm not a high voltage zener fan....noisy, and considerable voltage drift with temperature. I tried zeners in my tube amp and soon abandoned them.....
Go by the stabiliser circuit..... Frank posted.... using OA'S etc......great blue glow and discussion piece. Remember, tempting may it be; don't shunt a stabiliser tube with a medium uF capacitor or it will oscillate.
rich |
hi rich, at the end of the day the amp will have VR tubes for regulation. it is just that i wanted to find out if the circuit is really ****, which turned out not so.
| quote: | Originally posted by Gabevee
As for the 125ESE transformers, they do sound OK. The 300 B makes them go down to 40 Hz!
But when you get a pair of the Hammond 1627SE's look out! Those bad boys sound terrific!
Happy listening!
Gabe |
hi gabe, i originally wanted to use Tango U-808 for this, but i changed my mind and will stick to the Hammond and probably use 1628SE for a 5K primary. bad boys - and big too!
| quote: | Originally posted by burnedfingers
I ended up using 2) 6SN7's for the front end and I still can't get more than about 3watts RMS into a 8ohm load with a 1K sine wave. I have plenty of drive voltage to the 807 so I am scratching my head on this.
I don't know how people get 12watts RMS out of one of these.
J |
hi joe, i can't wait to measure power output on mine, which i have no idea yet. however, it seemed to be louder than my JE Labs SRPP 2A3 but not louder than my JE Labs 300B DX - in all cases no active preamp was used just a passive attenuator with a DACT in it. |
|
|
| Gary Gerhart |
Hello,
New guy here...
Just thought I'd toss my opinion into the mix...
A few things strike me right away when reading this thread.
First, the issue of output power. This is a Class A amp and. as such, efficiency of 35 - 40% is the best you're ever going to hope for. Class A is, by definition, grossly inefficient. So, if you have a B+ of 350v and an idle current of 40mA, the tube is dissipating, at idle, 14 watts (not counting the screen.) At Class A, you're never going to get more than around 4 watts this way. And distortion will be high, as the tube is just running too cold. An 807 is rated at 25 watts anode dissipation. Use it! 14 watts at idle is even cold for Class AB.
Using your updated schematic, if you simply change the cathode resistor (on the 807) to 220R, you should get around 8 watts before clipping with a 5K load. If you want more, you'll have to bump up the B+, but you'll soon reach a point of diminishing returns.
Next, the bypass cap on the 807 cathode resistor is just too small. You'll see a low freq roll-off much too early with the 47u. Bump it up to around 250u and you'll be down less than 2db at 20Hz.
The rest should work fine, but I, personally, would rather use a seperate screen supply or, at least, a voltage divider to derive the screen voltage. And I'd add a healthy filter cap at that node, then pull my driver B+ from there, rather than from the anode node.
Speaking of the power supply, along with the increased anode current comes, unfortunately, increased hum. The major problem with any SE tube amp. Because all of the anode current is *standing* in the primary of the OPT any ripple will pass through to the speakers. If you increase the idle current, you may well find that you need more filtering in the power supply. Your CLC section is good, but you may need to increase the 100u.
About the OPT. Those little Hammonds are not going to cut it if you make these changes which, IMO, you really need to make in order to really enjoy the 807s you've chosen to use. Though I've not personally used those trannies, I can tell you just by the physical size that you'll run into trouble on two fronts. The increased idle current is going to move the cores real close to magnetic saturation. Once the core is saturated, you're out of business. The 220R cathode resistor will have the 807 drawing around 60 ma at idle, and I'm farly sure that small Hammond core will be near saturation at that level. At any rate, you won't be able to send much program through the transformer before it saturates. The other problem, which is really the same problem, is the increase in size of the bypass cap on the 807 cathode resistor. This will extend the low freq range which will... tend to saturate the core of the OPT. And OPT saturation is a particularly ugly form of distortion, IMO.
As a cure, I can suggest two options.
I've heard great things about, but have not personally tried, the one electron SE OPT. It certainly looks good on paper, and I've spoken with people who have great success with them.
If it were my amp, I'd use the Mercury Magnetics AXIOM-SE-5K. I have used this one, and it is awesome.
Unfortunately, both of these transformer are quite large, and I doubt either of them would actually fit on your chassis, but I'm just going by the photo you posted.
Anyway, sorry for the intrusion, hope it helps at least a little... |
|
|
| arnoldc |
hi gary,
a most welcome "intrusion!"
i will certainly do these as my next steps moving forward, and i'll post the results here.
i checked the mercury magnetics web site and their prices are reasonable, those transformers sure look BIG!
the chassis is just a prototype so i wouldn't mind having bigger iron. :)
thank you very much on your recommendations!! |
|
|
| burnedfingers |
The shunt regulated screen supply with the VR tubes works great!
I junked out an old organ amp today :smash: :smash: :smash: it lent its chassis, transformer, and filter caps to the project. The screen supply stayed at 250-251 volts and I set the B+ at about 430. I had the capability to hit 454 volts but didn't like the sound of the amp at the higher voltage. :smash: |
|
|
| fdegrove |
Hi,
Frank knows, Joe.....;)
Just kidding but I'm still glad you're happy with it...:)
They're simple to implement and to my mind make all the difference between the ordinary and the extraordinary...
Cheers, ;) |
|
|
| arnoldc |
| hey joe, how much power do you get at 434V B+ ? |
|
|
| burnedfingers |
arnoldc
Quote:
hey joe, how much power do you get at 434V B+ ?
Well. I got a tad over 11watts into an 8ohm load with a 1K sine wave before it started to clip. It seems the VR circuit made a difference. I also used a larger transformer that I pulled from an old 6L6 PP amp.
I believe that I will lower the B+ feed to the plates to 400V and see what is sounds like there. It looks like the amp likes to see around 60mA to sound good. Surprising enough there is even some bass. The midrange detail is very good, somewhat scary when your not used to it.
J |
|
|
| jlsem |
| quote: | | Here's the shunt reg as per Gordon rankin: |
| quote: | | The shunt regulated screen supply with the VR tubes works great! |
Does the screen grid on an 807 really draw enough current for VR tubes to be effective? I hadn't thought so and am wondering if this improvement is psychological.
John |
|
|
| arnoldc |
| quote: | Originally posted by burnedfingers
The midrange detail is very good, somewhat scary when your not used to it.
J [/B] |
So true! My friend who heard it using a Bob Brines designed floorstander with Fostex FE167 commented about how that midrange showed its character.
I'd be making mine 400V too... and my itchy fingers are telling me to use 5842...
:clown: |
|
|
| burnedfingers |
Quote:
Does the screen grid on an 807 really draw enough current for VR tubes to be effective? I hadn't thought so and am wondering if this improvement is psychological.
Well John the screens draw 5mA at idle and upwards of 14mA when the amp is pushed. As far as it being psychological, what can I say? Before the VR's I used a 10K resistor to a B+ supply of 250volts and managed to get less than 4 watts out of it. When you figure that I get 11 watts out now something has made a difference. You be the judge.
Quote:
I'd be making mine 400V too... and my itchy fingers are telling me to use 5842...
What the heck, give it a try. Save your operating points of what your using now for future reference. Let us know how it turns out.
I ended up using a 6SL7 and a 6SN7 for the front end. It puts out a sh*t load of drive voltage.
J
Nothing exceeds like excess. |
|
|
|
