|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
69.40.62.190
Yeah, another 845 question. Sorry, but this one is interesting. I have been trying to figure out why the grid volts are different than what they are supposed to be when everything else is right on the money.
I got to looking at the plate curves for the 845, and for the operating points of plate volts 430 and plate current 62mA, the grid volts look to be -39 or -40 on the plate curves, not the -51 given in the operating points from stereo sound.
Have I got that right? If I do, is the stated -51 volts for bias a mistake or did they tweak it some how? Is so, how would that be done?
What really makes it interesting is the -39 or -40 volts bias predicted by the plate curves is what I get when I measure it with all the other points right on the money.
Also, DanL came up with a nice little cap coupling schem for between the 6sn7 and the 845, so no more dc coupling and everything is right on for the operating points. I have been trying different flavors of coupling caps and it is really starting to sound good.
Edits: 10/11/14 10/11/14Follow Ups:
The 845 plate curves show about -51 volts grid bias for a plate to cathode voltage of 430 and 62ma. (small red dot)There will be differences from tube to tube and the filament voltage will effect this as well.
If your filament voltage is low then the tube will not want to conduct as much and you will not need as much negative grid voltage (bias) to keep the plate current at 62ma.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 10/12/14
I now get what you have when looking at the plate curves. I will be curious to see what I measure when the filament voltages are correct, but since the current and plate volts are good I will let it be whatever it is, as you and Jim and others have suggested.
Thanks!
It sounds like you pretty much got it all figured out now.
I'm happy for you.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I had better take another look at my plate curves, because my dot ended up on the left side of the -50 line.
Thanks!
Vinnie
The low filament voltage will effect the bias point.
DanL
Aha! Suspicions confirmed. I Will get that up to par when I get a chance and see what that does for it.
Thanks!
"the -39 or -40 volts bias predicted by the plate curves is what I get when I measure it "
Dang, that's funny, Vinnie! So, your setup agrees with the published curves. Someone else's didn't, so now you're wondering if you did something wrong. Hilarious. I'm still lafin'. (Not at you, of course.)
BTW, I take it from one of your other posts in this thread, you've discovered that coupling caps aren't the satanic envoys you've been told. I was going to suggest that earlier (you know how I feel about direct plate-to-grid coupling), but my flame suit's getting a little frayed lately.
Keep up the good work!
--------------------------
Buy Chinese. Bury freedom.
It actually sounds as good as dc I think. Plus I get all the fun of cap rolling! So far a 0.2 uf k40-y9 is my favorite.
If what DanL says is correct in the above post, I may get closer yet. The filament voltage is down about a volt. I have to snag another 10v filament tranny to stick on the second channel. It will be interesting to see what happens.
Sure the biasing is easier with cap coupling, but what sounds best?
Keep adding up design compromises and you get a mid-fi result.
I think that is where most of us end up - in the mid-fi class due to the willingness to accept compromises in our audio systems.
Few really use the best parts or circuit implementation to be considered Ultra-Fi builders of SE tube amps.
Vinnie
If you want to step it up ...
Get a pair of grid chokes.
You will hear a difference.
DanL
Chip
Last paragraph of OT ...
Also, DanL came up with a nice little cap coupling schem for between the 6sn7 and the 845, so no more dc coupling and everything is right on for the operating points. I have been trying different flavors of coupling caps and it is really starting to sound good.
DanL
Never used them. How do they hook up and what do they do?
Vinnie
They replace the 100K grid leak resistor.
They provide a low DC impedance so
the 845 has a more stable bias.
They provide a high AC impedance so
the driver tube has lower distortion.
A Win-Win situation.
DanL
What do you recommend for a choke?
Vinnie
MagneQuest has the best as Tre' stated.
(Lowest Capacitance)
I use the BCP-16 Ni (Nickel/M4 Pinstripe).
Better performance than the M4 while
without the V restriction of the HN -
the High Nickel version.
2000+ Henries
3850 DCR
DanL
"replace the 100K grid leak resistor."
I don't mean to nit-pick, but that's not a grid leak resistor. I think "grid bias resistor" is the correct term, even when it connects the grid to ground.
--------------------------
Buy Chinese. Bury freedom.
TK
If it was a fixed bias I would call it that
but since it is only referencing ground - NO.
The bias is cathode supplied, not grid supplied.
But then again many things are totally illogical
in their nomenclature.
DanL
Now this is just my opinion and the way I look at this;I think a grid to ground resistor should be called a "grid resistor" unless the value of the grid resistor, in conjunction with the grid leakage current (not grid current), is being used to bias the grid positively as is sometimes done with very low plate voltage when the only way to get any plate current flowing is to bias the grid a little positive.
I think they did that is some old car radios and in that case the term "grid leak resistor" should be used.
Just my 2 cents.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 10/12/14
"I think a grid to ground resistor should be called a "grid resistor" "
That works for me. I just don't think we should refer to it as a grid leak resistor unless it actually serves that purpose. Too much chance for misunderstanding, not to mention the confusion it might cause a newcomer.
--------------------------
Buy Chinese. Bury freedom.
"A Win-Win situation."
Yes it is as long as the grid choke has low shunt capacitance and the driver stage can handle it.
Otherwise you will lose highs.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
So what do you recommend for the choke?
MagneQuest BCP-16 GC
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Tre'
The 6SN7 should drive any grid chokes.
The only posible exception might be
the EBAY TubeLab ones.
But they don't publish their specs.
DanL
When I ask Kevin what the winding capacitance of the LL1670 is he answered 223pF.
This would be added to the Miller of the tube.
On the other hand Mike said the measured capacitance of the BCP-16 GC is only 15pF.
I should buy a pair.
On the other hand my 5687 driver tube is CCS loaded and the 270k ohms grid resistor gives a pretty horizontal load line and I don't have enough gain in my system to ever overload the output tube.
So, as with most things, the particular circumstance dictates what's needed and what's not.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Sounds good. I will give it a try first chance I get.
So far I can not hear any difference that is significant. In some ways though it does sound better than the dc version. But then I am using a small cap of high quality.
When I was doing a lot of ARTA measurements, I found that the biasing, plate loading and tube rolling were a few things that seemed to affect the % of different harmonics at the output of the amp.
Most of the amps I have are also cap coupled SE but I did not get too much into cap rolling and seeing if that affected the harmonic profile of the circuit measurements to any degree.
I used mostly Russian oil caps or NOS film types for cap coupling.
As for reference to anyone else contemplating low voltage 845's, here's a post from Paul Joppa way back in 2005 suggesting operating points with a 5K load.
Thanks for the link, that's the first time I have seen that one. If you follow the links back there is another set of operating points as wellVp-k Vbias Iplate Pout
400 46 53mA 4.6
450 52 60 6.0
500 57 67 7.5
550 64 74 9.3
600 69 80 11.0These seem closer to the ones that Mike Lafevre posted some time back.
Edits: 10/11/14
Vinnie,
Please look at this excerpt from my FAQ post on the Tube Asylum:
"Keep one last thing in mind - often you see a voltage "spec" on a schematic at the control grid of a tube. DO NOT TRY TO SET THE VOLTAGE AT THE GRID TO THAT SPEC! The negative voltage at the tube grid should be whatever value is required to set the tube's cathode current to spec. If the tube is passing the correct amount of current and the voltage at the grid is -20 volts then the hell with the "spec" of -22.5 volts or whatever it is. Making the grid voltage more negative with respect to the tube's cathode is the equivalent of stepping down harder on a garden hose to restrict the flow of water. And if you let up the pressure on the hose and more water flows - that's the same thing that happens if the tube grid is less negative!!"
Variance from tube to tube is why this all occurs. Not every tube passes the same amount of current even if all else in the circuit is identical.
Was I reading the plate curves right though Jim? Do they actually predict about -40 vdc for the bias with plate current of 62mA and plate volts of 430 vdc, or am I doing that wrong?
Edits: 10/11/14
Let this go Vinnie, you're just confusing yourself. There is often a LARGE variation in plate and/or cathode current between tubes. So the curves may show -40 but that's only true for a few tubes of the type. Due to internal construction variances & variances of heater voltage the -40 might need to be -32 on the next tube or -51 on the next after that and so on.
The curves only apply to a "bogey" tube under very specific circumstances. The typical tube in a typical amp often shows significant variances even with identical circuit characteristics - because the tubes themselves often perform far differently than the bogey tube does.
You misunderstand Jim. I am just trying to figure out if I am doing it correctly. I am quite happy to go with your advice on not trying to mess with the grid volts, but I do want to know if I am reading the predicted value right for future use.
Besides, if I am it means the readings that I am getting for bias matches the predicted and the bias listed in the operating posts is much different for some reason. I would like to know that if it is true.
Edits: 10/12/14
Vinnie,
The exact grid voltage is essentially meaningless. Only in the most general and loosely defined terms does it have any meaning at all. The tolerances on grid voltage are quite loose so the curves only approximate what you'll find in real world situations. That is because the curves are only correct if the tube behaves exactly as designed.
You can look on a plate curve diagram and if you know the current and the voltages then the diagram will tell you what the grid voltage should be - as long as the tube is built EXACTLY to the spec - which they almost never are.
Can I suggest a little research on how tubes work is likely to make this much more clear? Your posts have changed the verbiage some, but essentially you are asking the same question over and over.
Should the grid voltage on the curve diagram match the tube in your project? In a perfect world maybe - but in the real world the tube's tolerances render an exact number meaningless. The next tube in the EXACT SAME circuit will behave differently.
JimThe link below is the answer I was looking for; you just kept misreading the question.
I was reading the plate curves incorrectly, that's what Tre' showed me in his post and that was what I was asking you.
Edits: 10/13/14 10/13/14
Vinnie,
What I'm trying to get across is that the curves Tre' posted are only accurate for a very few tubes. The VAST majority of 845s will not end up with -51 on the grid even with 430 plate volts and a current of 62 ma.
Just keep that in mind!
We have probably beat this to death, so "Roger on that". : )
Thanks Jim. That was brought up in the original thread a few days ago, but it was more like "if the plate current and plate voltage are OK that is more important than the grid voltage". What you are saying is similar but different. I was curious why I could not get it closer; now I understand.
By the way, how do I find that FAQ post of yours? It sounds like there is more in it that I need to read.
Hi Vinnie,
Go to the Tubes Asylum and click on the FAQ link at the top of the page. The post is entitled "Tube Problems" That Aren't Tube Problems" . Or click the link below...
Thanks Jim!
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: