|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
80.60.8.110
In Reply to: RE: Fixed bias problem, and no, it is not the tubes. posted by DAK on January 11, 2017 at 18:00:49
The bias transformer should not get hot. It has to deliver only very low power.
Edits: 01/12/17 01/12/17Follow Ups:
I put in a 10K 3/4W DALE rn65 resistor and i monitored the amp closely for over an hour and the bias voltage at the new resistor and at the power tube cathode both stayed unchanged over that time. I am going to pop in some good tubes and check the voltages again. Thanks for all the help guys. regards, Dak
Well, i put in some KT90s that test really strong and that i have used many hours with no problems, and the bias current started to run away again. The bias voltage also was dropping so there is definitely something going on. The problem is affecting both channels so it must be something from the filament trans to the 10K resistor, since both circuits at the pot seem to be affected equally. I think i will try a new filament trans.
I was checking my filament transformers and i have one that is 6.3v @ .5A. Is this trans ok for my circuit? Which would require enough current to bias 2 x kt150 tubes.
You may have one that is leaky.. If you tried everything else,you may have some DC on G1.
"For every complex problem there is an answer that is clear, simple, and wrong" H. L. Mencken
Can you post a schematic of the bias supply including adjustment pots and grid resistors?Thanks
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 01/12/17
Hi Tre, i could not find the schematic in my files but it is a very basic type. The filament tx is reversed wired to a .8A 6.3v tap, the 120v is FWBR and filtered by a choke about 25H x 715dcr. this voltage is filtered by 50uf x 120v then dropped by 5K 3/4w mf then 50uf x 120v then dropped by 10k 3/4w mf then 100uf x 100v to the 50k AB 2W carbon pot. The wiper is 50k 1/2w cc connected to the power tube grid stopper. The wiper also has a 50uf x 63v connected to it.
Well, i did not have a single 50uf x 120v cap so i series connected 2 63v x 100uf caps, and not remembering that the metal can is "hot" sometimes they were sitting on the chassis. I guess i better change them out. thanks for the help! cheers, Dak
"The wiper is 50k 1/2w cc connected to the power tube grid stopper"Did you mean the wiper of the pot connects to the bottom of the grid resistor?
BTW I think you said there is -60 volts feeding the pots?
two 50k ohm pots in parallel to ground is a 25k ohm load.
60/25000=2.4ma. of current. That should be the total load current on the bias supply as long as the output tubes are never driven to grid current (overload).
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 01/13/17
I changed them out and the amp worked fine at the shop for almost an hour everything stable with nice audio. Brought the amp home hooked it up was stable for about 10 minutes then the kt90 bias current started to climb and i shut it down when the meter was at 2.0.
So i unhooked the amp and put my 2 chassis amp in place with the same kt90 power tubes. this amp is the same circuit with fixed bias. I hooked everything up and turned it on. Adjusted the bias to .90vdc or 90ma and everything is good. No bias drift the meter is showing no problems. The sound is very nice.
Regarding you picture. That is exactly how i have it with the grid resistor being 50k. The actual bias voltage before the pot is -51vdc. The bad amp is SEUL with Edcor outputs the 2 chassis amp is SE pentode with Electraprint outputs.
I dunno, the fact that it seems to work at the shop and not at home is really pissing me off. cheers.
Edits: 01/13/17
You only need voltage. You can bias these tubes with less than 5mA. Sounds like you have a bad cap in the bias circuit that is drawing current. That or a bad transformer.
Edits: 01/12/17 01/12/17
Ok, so the .5A 6.3vac filament trans should be perfectly fine for the job?
I still don't understand why it worked in my shop for over an hour with solid cathode current on the output tubes of 75ma and the bias voltage was rock steady as well at -52vdc measured at the last dropping resistor. I took out the "test tubes" which were a mismatched duo of a GE 6550A and a Russian kt88. Plugged in the kt90s and after about a 5 minutes the current started to steadily climb while bias voltage steadily lowered. I turned it off, put in the old testing tubes and this time they could not hold bias worth a darn.
A faulty diode (it may become 'open circuit' or a sort of non linear resistance) in the full wave rectifier of the bias supply might turn it into a half wave rectifier.
As a consequence DC current arises that might overheat the bias transformer.
If KT90 tubes are brand new I would also put a temporary cathode resistor durung the initial 200-300 hours ageing (to be eventually removed later on).
During the initial burn in some excess air molecules may be trapped inside the tube, excessive grid current results and getter action takes a significant time to clean up them.
"In bocca al lupo!"
ecc230
The bias voltage starts out normally and at the shop it was running normally for over an hour. That doesn't seem to indicate a problem in the fwbr. I even changed the battery on my Fluke meter in case that was the problem. I am going to test again now that it has been sitting for over 12 hours.
The KT90 are used but i will remember your tip about burning in new ones. thank you for the tips.
Schematic will help. Almost sounds like you have an induced instability or parasitic oscillations.
What kind of grid stopper and screen resistors are you using?
I am beginning to believe that with nothing else left to rule out besides the transformers, it might be an oscillation that is inaudible. It is probably easier to add some resistors to tame this but I would like to discuss off forum for now. cheers. Dak
I really don't see that as i have made at least 3 of these amps. Here is the basic outline http://rh-amps.blogspot.com/2013/05/rh-universal-v2-totally-universal.html
My amp differs with a 6sn7 driver tube, and fixed bias. I have made the fixed bias version 2 other times with no problems at all.
well...disconnect the finals from the rest of the amp. If it misbehaves still...
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
It took roughly the same amount of time for the power tubes to start drawing excess current. Does that rule out the driver tube circuit as the source of the problem? I guess i should replace the bias power trans?.
take the coupling cap off at the power tube's grid. You'll need a couple meters with their leads clipped to a few places. The power tubes current sense, and some of the bias supply. Then fire it and you should find the problem.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
.
I offer an example; hook at least one to the bias supply. Absolutely have one on the final's cathode sense resistor; that is your primary indicator, yes? So, if voltage on the bias supply remains constant, and cathode current climbs, what does that tell you?
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Right, thanks, i will do those checks.
"Right" is not usually an acceptable answer to an essay question...:)
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
I had just crawled out of bed, and these days, i mean crawled, and dashed off an answer. Nowadays, dash is just a concept. I will give an update after i get to the shop and try some of the suggestions.
Edits: 01/16/17
What I was hoping to see was an answer to the last question. Not jacking you up( or at least trying not to). Attempting to gage level of understanding of a proposed scenario.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
It doesn't rule out the coupling caps.
Not that I think that's your problem.
I still think it's filter caps in the negative supply.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
That would seem to indicate that the bias supply caps are good.
I guess you're down to environment induced oscillation.Grid stoppers and ferrite beads?
Maybe a bad PT.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 01/15/17
I can put a grid stopper no problem but where would i put the ferrite beads?
I would start with the large snap on type on the power cord and audio input cable.
Then the tiny donut type around the wires leading to the grids.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I got a snap on the power cord already. But i can put the bead type on the inputs. Do you think a grid stop resistor is warranted on the input/driver tube? The amount of gain from the driver tube is already on the low side since the amp is only 2 stages. thanx tre.
If the amp behaves in a different location then we are assuming that you have a lot of RF at home so, yes. A grid stop resistor on the input tube might be just what you need.
It's easy enough to try.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
The amp was stable for about 10 minutes then slowly the cathode current of the 6550 started to tick up, and then it picked up speed and i shut it down. I also put in a grid stopper on the input/driver tube. I put the ferrite bead on the wire lead to the grid stopper. it is short not even 2". Maybe i should go with a shielded input wire? The run is short as i said.
Edits: 01/16/17
DAK
Just disconnect the coupling cap going to output tube and run it.You only have to disconnect the cap on one side and you can't play music but you can run the amp at idle to see if it drifts.It may not be your problem but it's quick way to find out.
"For every complex problem there is an answer that is clear, simple, and wrong" H. L. Mencken
If one is bad i can see that as a possibility but both new caps being bad does seem unlikely. Even if 1 cap is bad can it pull the bias voltage down equally in both power tubes?
Even if 1 cap is bad can it pull the bias voltage down equally in both power tubes?Manufacturers do things in runs.If there is an issue with those caps,being they came from the same run no doubt,that can be an issue for both.
It's very simple to do that test and get it out of way.That may not be your problem but newer oil caps have had issues at times,including the Jensens.There may be cases where the oil has become conductive for whatever reason. You have built these amps several times and this is the amp you are having trouble with. I would try the caps. Just disconnect one side of each cap and run it and that way you will know.
"For every complex problem there is an answer that is clear, simple, and wrong" H. L. Mencken
Edits: 01/16/17
Since the amp seems to be "unstable" only at home seems meaningful. At home i have wifi, several computers hooked up to it and also a cellphone signal booster all in the same room.
The amp is a simple 2 stage SEUL, 6SN7 to KT90/120. in the design there is no grid stopper on the input/driver tube and the grid stopper for the kt90 is 240R. I have made 4 of these amps with no problems with instability. In this version the 6sn7 is biased with 2 LEDs instead of a 380 ohm resistor. This increased the gain of the tube because it made the amp louder compared to the carbon comp resistor.
Edits: 01/14/17
And it slowly declined after about 5 minutes. The coupling caps are 2 brand new Mundorf Silver and Gold in oil. The voltage rating on those are 800vdc so i seriously doubt that it is defective. The 10R resistor is a 2 watt unit and i did check it at the same time and it was not even warm, so again, doesnt appear to be a problem.
There is your problem. You have to check the bias voltage not only at the end. Where you find the decline, there is your bad part. Never assume good parts cannot be bad.
When I read the other answers you gave in this thread, I realized that it could not be the things I mentioned, because both sides (left and right) behave the same.
So it has to be something different.
Then I read your answer that the bias transformers gets hot
And that should not happen.
So I would replace the transformer.
Hi AD, i have 2 tx in the circuit, the filament trans and the choke. The filament tx gets warm and hot at about 30 minutes of use (i noticed the heat during the first few times i powered up the amp and now i notice that it gets warm after about 5 minutes) the choke is cool. My feeling is i should try a different fil trans.
AD has a point. As the transformer heats up the DCR is drifting. Perhaps check your connections first as you said it was fine in the shop.
That the bias transformer heats up can be a problem of the transformer, but could also bet hat one of the capacitors in the bias supply is not good (internal leak) so the bias transformers has to deliver some current.
Edits: 01/12/17
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: