HomeBrew Troublshooting Help

[chunkitup]

It's a pity you don't have a scope, to see what that "hiss" really is or looks like.
After all, some RF oscillation, (completely inaudible by itself), may have audible artifacts, which could be confused with real hiss.
Just to check two more points:
1) pull C1, its insulation might be flaky, remember it carries around 150V across it. Yet it might "measure" good.
2) solder a 100pF ceramic grid to cathode in V8, on the socket itself.
3) solder a 1M resistor, straight from V8's grid to R4's grounding point.
4) pull V8, tighten the socket connection with a sewing needle, clean it with whatever you are used to.
5) throw your black cat out of your house, or, at least deport it until you solve this problem.
Good luck.

I actually believe it was RF oscillations because I put plate snubbers on the first and second stages as well as the third stage and it made the amp very quiet. In fact the owner was quite pleased with how quiet. I am just mad that the tube depot said the TungSol 5881's were alright with the 430v on the plates. Perhaps it took only a few months to wear them out.
Also, I did just buy a Tektronix 2246 scope with 10x probes. I need to catch up on proper use since my last electronics lab class was in 1991. Can you give me some do's and dont's? Refresh my memory.

[J M Fahey]

Hi.
Maybe you have a problem with old TungSol's, but *this* problem hiss, that disappears pulling the first or second preamp tubes, does not come from them.
Maybe somebody can point to a basic audio 'scope use page; being 10:22 Saturday night I'm not exactly in the mood to write that now.
Basicaly you want to inject , say, 50 mV @ 1KHz into your input jack; measure and *see* around 2.5V (7Vpp) on the first plate, (use the AC position or hook a .1uFx600V polyester or polypropilene in series, you *don't* want the around 250VDC present there to swamp your AC signal), and after checking it's a nice clean sinewave, which you can also hear through your speaker (rise the volume to some real-life sound level, that's to say, annoying) *and then*, touching nothing, unplug your signal jack.
If you see a fuzzy, ghostly signal, which is really a lot of thin sinewaves, so close to each other that you see them as a splotch of light, that's your RF. It is confirmed by switching the frequency or sweep control of the scope higher and higher, until, when you are seweeping into yhe hundreds of kilohertz you will see those very high frequency sinewaves, most certainly quite distorted, skewed, ugly.
The other dreary possibility will be hiss , which looks like grass on the "floor" of the signal sweep, remember now we have no sinewave applied.
If some capacitor's insulation is failing, you will see random spikes.
Why did I tell you to rise the volume to rehearsal levels?
Because then you set the amplifier gain realistically.
If you set it too low, you will hear no noise and pronounce the patient cured.
If you set it too high, you'll get crazy chasing real normal hiss/hum.
Good luck.
Note: concerning Murphy's Law, very often the scope probe or cable capacitance tames the oscillation and you see nothing, yet when you pull it off it reappears, (maddening), that's why I asked you to use the amp+speaker as a backup system.
Besides, watching the screen and hearing simultaneously is *great* training.

[Gtr_tech]

http://frank.pocnet.net/sheets/127/5/5881.pdf

Thats what they were trying to recreate. Look at teh voltage ratings.

I think since the 5881 moniker has been used fast and loose by Sovwreck, ppl have forgotten (or just never knew) that it is actually a *derated* 6L6GC type. Actually closer to a 6L6GB.

[macpgh]

1 eyed jacks

think about the input jack. Its the mechanical aspect thats been used. Could be a bad one or a bad ground or bad material in construction.

Think about soldering technique......some spatter may have found its way into the volume pot on that channel. You can spray it but it will probably come back & its best to change them out.

I use a jig where the pot is turned to have the lugs down at 6 o'clock to solder & then its placed into circuit/chassis

BTW 430 volts is still within 20% tolerance for a 400 volt tube by book spec's

TUNG SOL's were the ultimate premium tube of the day & 5881 is a military spec tube

[chunkitup]

Well, I figured out the problem. I built my own bias probe and low and behold the 1ohm resistor was .5ohm's. When I thought I had 40mA I really had 80mA. I am surprised the tubes lasted as long as they did. I will learn from this mistake. Never believe that when it says 1 ohm it is 1 ohm, always measure!

[tubeswell]

Well, I figured out the problem. I built my own bias probe and low and behold the 1ohm resistor was .5ohm's. When I thought I had 40mA I really had 80mA. I am surprised the tubes lasted as long as they did. I will learn from this mistake. Never believe that when it says 1 ohm it is 1 ohm, always measure!

FWIW when I install the 1R resistors, I use 2 x 2R 1% (the little 0.6W metal film ones) in parallel for each '1R' (and I measure them first). This theoretically increases their chances of being 1R or damn close to it. 2CW

[J M Fahey]

Now it got interesting: ¿how do you measure (with precision) a 1 ohm resistor?
Thanks.

[chunkitup]

With my DMM.

[tubeswell]

Now it got interesting: ¿how do you measure (with precision) a 1 ohm resistor?
Thanks.

Measure the DC resistance of your meter leads on your DMM. (it might be something like 0.5 - 0.8). Then measure the resistor (you will probably get something like 1.5- 1.8 accordingly) - i.e.; there should be 1R difference (if your meter is good enough). If you use 2 x 2R 1% resistors (or even better 4 x 4R 1% etc) in parallel, you will get fairly close to 1R anyway (the bigger resistors are 'easier' to get a measurement on). You can always calibrate it with another biasing method (like the transformer shunt method), but I don't bother with that, I just use 1% MF resistors

[chunkitup]

If it isnt one thing it is another. Why dont I just disconnect the ground and put the DMM in line and set the meter to the mA scale?

[J M Fahey]

Hi.
I asked because regular DMMs are *not* trusty measuring such low resistors; of course chunkitup might have an expensive special 4-wire one or made a precision 100mA or 1A power supply.
I use the last one, and also have a Leader RCL bridge.
I make lots of SS amps and my own 0.1 ohm precision resistors to use there, so I had to tool up.
I was intrigued by the 100% difference in resistor values (1 ohms -> 0.5 ohms) ; because what I usually see is them being on the mark, even regular carbon film ones rated 5% usually measure within 1% to 2%.
As to using the meter in series with the cathode: it will certainly be an exact measurement, but it is *VERY* dangerous, you can easily have hundreds of volts there or, to be more precise, the tube will try to draw the current it needs through you, if your meter point slips or whatever.
That's why the cathode resistor method is preferred.
Thanks.

[chunkitup]

You are most correct to wonder about the 100% difference from 1 ohm to .5 ohm.....and it goes to show how forgetful I really can be. I have to shake my head at how stupid I can be. I thought I was measuring a 1 ohm resistor on the bias probe but there IS no resistor. The probe puts the meter in series with the cathode to ground. I was measuring the resistance of the probe.
As far as the TungSol 5881's max plate voltage is concerned here is the response I got from the tube store:

It is asper Otto, the head tech at New Sensor, the maker of this tube. It's also found in any generic 5881 data sheet. Tube Depot should really know this stuff.

And here is the data sheet:

http://lib.store.yahoo.net/lib/thetubestore/ts5881.pdf

which shows a voltage of 360v and if you figure in the 20% it would be a max of 432v and I had a voltage of atleast 430. So just to be safe I put in TungSol 6L6's. The amp sounds great.