Bottlehead Forum
Bottlehead Kits => Legacy Kit Products => Stereomour => Topic started by: Alonzo on May 25, 2015, 11:37:23 AM
-
Just needed to keep my hands busy so got this done in time to spin tunes while catching up with some other vets.
-
This 2A3 Stereomour had some hiss in my horns. To correct it I ordered a variety of 431's to try to get it lower. After 3 try's I went with 665 ohm cathode resistors (I think that's the correct term for the resistor from 3 and 8 to ground on the 12AT7). Still had hiss but after a dozen or so tube swaps I accept it, it is what it is.
Somewhere along the way I broke something, turning on the amp today after work gave a couple of loud tics and a loud squeal that got louder and louder till I shut down the amp. Swapped power tubes, same result. Swapped 12AT7's same result.
I'm assuming it's oscillation. It's new to me, I've had motorboating, this sounds different. Anyone have a "readers digest" version of what oscillation is/does to the amp?
Going to re-wet all the new connections I've made since the last working version of the amp and hope that fixes the problem. If not then will try the chopstick search for bad connections. The amp worked well before the resistor changes but I'll check everything.
Another question, this amp is connected to a Smash, which has a 45 amp also connected. When the 2A3 is powered up (and squealing) it dampens the 45 amp volume or the Smash volume. I have to adjust the Smash a couple of clicks back and forth to get it back to my normal listening level. What is the feedback from the 2A3 that is causing the Smash tubes to be dampened?
Thanks in advance for the help,
Alonzo
-
I tried the 431 experiment in my S.E.X. amp and gave up on it due to hiss. I also bought half a dozen different variants but none of them were quiet enough. What is the bias level you are aiming for as there might be an LED option out there you could try.
-
Just following the instructions from Paul J, the resistor is to drop the voltage 2.5 volts. He even mentions led's in the forum response.
-
I see a lot of modifications, which means I don't know what the current circuit is, which means I can't make reliable guesses.
Oscillation is a broad term which can have many causes and many "features" - there is no condensed version.
Since there were no changes immediately before the squealing problem, we'll have to start with a blank slate. We can start with what changes were made? You said the noise gets louder until you shut it down, so there is some risk in running it to determine the voltages. If I know the changes I might guess whether to make changes before further testing. At least we know it worked before this last problem!
-
The last working configuration was as a 2A3 amp with 12at7 input with C4S (still using the 431). No selector switch and 100K resistors instead of a pot. M4 grid chokes instead of the 249K grid resistor. Ten turn 50 ohm hum pots, 15uF caps across pots. Dowdy plate chokes instead or the PC-3, BH-5 output instead of the OT-2, 30uF parafeed cap (.3uF bypass) instead of the 3.3uF original cap. In the power supply used the PC-3 at 270 ohms instead of the parallel 1200&360 resistors, Cree diodes instead of the UF4007. Ran this way with 2A3-40 or 2A3 meshplates, different 12AT7's for a month or so.
Changes I made leading up to the squealing;
4 replacements of LM431's. Each with a startup, play around 4 minutes (nickel hum disperses) then check for hiss.
Added 2uF PS bypass caps from HV+ to ground, checked for hiss.
Cut black wire from B3 and B8 at Kreg, added 665 ohm resistor from wire to ground, wire doubled on itself and twisted twice. Started up, a little static pop and crackle, no changes to hiss, played a couple of hours.
Came in the next day, on startup, couple of pops/crackle and the squeal starts. Fuse is good, swapped power and input tubes with no changes.
So that's where I'm at today. Thanks in advance for any/all help in diagnosing and repairs, but especially for explanations, its another opportunity to learn about amps.
-
You'd need to post your voltages for us to have any idea what's actually going on. Since your build is so customized, you'd want to list them as driver plate, driver cathode, driver grid, output filament, output plate, and output grid.
-PB
-
Paul,
Are you Ok with voltages without the power tubes? I've plenty of 12at7's so those can be sacrificed, I'm just afraid of blowing out some 2A3's because the squeal never peaked before I shut down the amp. Halted Electronics is around the corner, I'll see if I can find some cheap 2A3's tomorrow.
-
Without 2A3's in the amp, the voltage measurements aren't necessarily going to be super informative.
-PB
-
Hi Paul,
Here's the voltages; using the Stereomour labeling where I can and adding part names where I know: all VDC
Output filaments:
A1 61 C1 63 A4 61 C4 63
output grids:
A3 0 C3 0
output plate:
A2 376 C2 374
B+ Right (IB) 397
B+ Left (IA) 397
Driver Plate (OA) 264
Driver Plate (OB) 133
Terminals:
1 376 15 = 375
2 263 14 = 138
5 60 11 = 63
16 60
17 0
18 0
19 0
20 63
Alonzo
-
Thanks for the summary of differences and the voltages - that will help a lot.
The 2A3s have good voltages, indicating that the squeal, if present, is not causing overloads on the tubes.
The only voltage this looks bad is the driver plate (OB). This looks like there is a bias problem on that section of the 12AT7. Some possibilities:
1 - the resistance from the driver cathode on that side is much less than the target
2 - the above resistance is zero (a short to ground), AND that half of the 12AAT7 has very low emission
3 - the grid of that half of the 12AT7 is floating - possibly a broken grid stopper resistor? - possibly a bad 100K input resistor that replaced the pot.
My money is on 3, but I'm not putting much of it up.
-
Thanks for the info Paul. A little more reference points.
B8 and B9 were touching, I think from when I twisted the wire from B8.
Kreg reads .2 and .16 instead of 2.5
I'm going to re-wet the grid stoppers and 100K's and the Kreg connections. Then re-measure voltages.
-
Replaced driver tube, took out the 665 ohm resistors, connected 431 back at Kreg. Now OB is at 197 and Kreg reads 2.5 on both sides. Terminal 14 is still a little low, will confirm that lower connection is good and give it a speaker test tomorrow.
-
All's well. Sounds pretty good but I'm back to the original problem. May have to change speakers since I want to put some years on this amp. I will leave it somewhat stock, well in my version of stock at least.
Thanks for the help!
-
Are the speakers very sensitive?
-
I think they are 93db rated. Just going to have to learn to live with the hiss or see if the Smash is the component injecting the hiss vs the amps.
-
OK, so the resistors gave a squeal which was cured by restoring the 431. But we are using resistors in Stereomour II with no problems, leading me to suspect the wire loop even though most of it is twisted.
It would be valuable to know whether the 431 is actually the source of the hiss. The resistor experiment was intended to answer that. The impedance of the 431 is so low that bypassing it with a capacitor won't have much effect. An alternative is to remove the preamp (use shorting plugs) from the input of the Stereomour. Perhaps you have already done this - I didn't see it in a quick re-read of the thread.
If the hiss is audible on 93dB speakers, it might be measurable as an output voltage on the speaker terminals. That would make comparisons between configurations more accurate and useful.
-
You should be able to beat this. I have ~95db speakers and have no notable hiss with 45, 2A3 or 300B. a little hum, but no hiss...John
-
Paul,
More sample points. The original LM431's that came with the kit had hiss, I don't know the kit purchase date, I bought the kit from Jim Redman (I'll search my bin for those and get you the type). I tried the following ones, 2 tests from each type;
Fairchild LM431ACZ, STMicro TL431ACZ, all 5 tries hissed but the STMicro's were the best tried in the amp. I have TL431AILPG from Digikey that are ON Semiconductor but manufactured in that country that copies things a lot. I will give these a try the next time I'm under the hood :)
I'll give it a check with shorting plugs and also measure the output volts with the current 431's and with the next pair.
John, I know its there, eventually I'll get it figured out. I may have to accept it though, I've added a lot of iron and caps, the baseline noise floor may be higher than I expect.
Thanks,
Alonzo
-
I've had this same problem about a month ago. My set-up is also a Smash and Stereomour 45. It turned out it came from the Smash. I changed the Smash tubes and the "squeal" went away. The next day, I put the old tubes back and it played fine again. After about an hour, the squeal came back and this time I held the tubes with my hands and the squeal slowly went away. Didn't have same problem ever since.
-
Will be rebuilding the C4S board this weekend, messed up something while changing 431's and one channel of LED's blew. It was an awesome pop, but didn't blow the fuse. So somewhere I have a problem that caused too high a voltage maybe. Going to the ON Semiconductor 431's and will post the board voltages and the output voltages.
-
An update in the search.
Both Stereomours now have STMicro TL431ACZ's and both still have a hiss that is apparent with no source playing and in the quiet passages or single instrument music from about 5 feet from the speakers. Each amp measures less than 0.020V. It can be adjusted to 0.000 for about half a turn of the hum pot (10 turn pots) then jumps back up to 0.020.
Next steps;
Made up some shorting plugs so will test/measure with them later.
Tube roll the 12AT7's
Swap out the 431's in the Smash for STMicro's
purchase some extra boards so all can be changed at once
Try the TL431AILPG in pre and amps
-
Just had a read of my own thread to remind myself, http://bottlehead.com/smf/index.php?topic=5316.msg76160#msg76160 (http://bottlehead.com/smf/index.php?topic=5316.msg76160#msg76160) in my application i found the On Semi TL431AILPG was the quietest of the lot (tried half a dozen), but ultimately none of them were quiet enough. Keep us updated as i'm curious to know where you get with this...
-
Yep, will do. Trying the On Semi's next but my boards are crappy now from all the desoldering and installs. Sent Eileen an email to see if I can buy a few more. In the mean time I can tube roll.
With all this, the amps do sound wonderful. It's just that nag that now that I know it's there I have to at least try to fix
-
Is the shunt regulator/active load printed circuit board for the Stereomour available to purchase? Haven't heard a response from Eileen via email, thought I'd post the question and call tomorrow. Is the 4.4 version for the SEX the same or does it have extra traces?
-
Eileen is out of the office. Send another e-mail later in the week after she has had a chance to go through her correspondence.
That is the correct board.
-
Ok thanks!
-
Just to finish this out. Currently all the TL431's in my system have been changed out. Received new SmashUp boards today and put them in with ON Semiconductor 431's. While assembling I ran both amps into my horns to test, only a very low hiss could be heard at about the 1" range from the speaker, with no music, drowned out with music playing. Finished the SmashUp and the hiss is back but not equal, left side is louder, and really obvious in the tweeters (duh). Threshold moved about a foot out from the speakers, with no music and with music you can hear it during quiet passages.
So it seems the Smash is amplifying the background hash I have. I'm about done with chasing it in the pre and amps, I'm going to work with maybe adapting one of my Fix's to filter it next, maybe pad down the tweeters more.
-
Finally fixed!!!! Replaced the Smash w/Smash-Up with a passive preamp. No haze, no hash, no hiss. Loving Mike's B7 Mexico series input transformers.
The ugly wiring job shown...