My FIRST tube amp (and it's working!)

[printer2]

Going through the 470 ohm resistor is no big deal, just gives better filtering.

Yeah it'll fly in a LOW powered amp, but you want the most filtering at the preamp stage, where noise will be amplified more.  Fender (and everybody else) did it right.

If grounding one side of the OT secondary killed the output, probably grounded the wrong side.  Some are grounded to case internally.

I figure if you have more voltage than you want there is no real downside with a resistor in line with the first position. But looking at the schematic I never bothered to look at the next stage 470 ohm is way too low a value, almost doesn't matter if it is there or not. Dropping less than a volt across it, I would go at least 10k.

[rankot]

Is it possible that I fried 100W 15" speaker with this little amp? As I already mentioned, it worked for a few seconds, then stopped at all. Secondary wasn't grounded at the time it happened. I grounded secondary after, but it didn't bring any good. Now I disconnected that speaker, connected it to my old SS amp and it seems it's dead What could happen to it?

So I connected 2W 4" speaker to this tube amp and it works, although sound is not really loud, but both amp and preamp respond to volume and tone controls.

[anotherjim]

Is it possible that I fried 100W 15" speaker with this little amp?

We'll hope not. It may have been a weakness in the speaker just waiting to fail. I'd continuity test all the way to the voice coil behind the cone if you can make contact with the wiring there.

[rankot]

I can see you have a complex layout More parts more work to achieve good tone, but  I admit very creative never thought about building one that way

All depends in the efficiency of the tranformers, lets say you have decent ones, industrially produced, so 85%.

14x3.5x0.85=41Watts

12x1.25x0.85= 12,75Watts

12.75/220=58mA

Which kind of rectifier are you using?

I had to use this back to back topology, because I couldn't find appropriate power amp PTF and I had few 220/12V transformers at hand. However, it worked well as preamp only PS, but it obviously lacks power for amplifier. So I will try it with a 50VA 220/12 toroidal TF I have.

Calculation in my case is as follows:
14x3.5x0.85=41Watts
preamp heater = (regulated to from 12V AC to 6.3 DC) 12*0.35=4.2W
amp heater (12V AC) = 12*0.3=3.6W
So availble power for second PTF which will generate B+ is 41-4.2-3.6=33.2W
If I need 68mA for both amp and preamp on B+, it means that I need 0.068*220/0.85=17.6W taken to the second PTF or 1.5A at 12V. It means that my 15VA PTF is quite insufficient.

[rankot]

Is it possible that I fried 100W 15" speaker with this little amp?

We'll hope not. It may have been a weakness in the speaker just waiting to fail. I'd continuity test all the way to the voice coil behind the cone if you can make contact with the wiring there.

I'll have to remove it and see what's up.

[rankot]

I can't access my 15" speaker's coil, but if I measure its leads, resistance of a coil is approx 20 ohm. Is it OK - speaker is declared as 8 Ohm? Measuring 2W 8 Ohm speaker gives 8 Ohm resistance.

I have also replaced secondary PTF with 220/12V 20VA, and the B+ is now 167V under load (270V before PCL86 heater reach working temperature). So it seems I need something bigger here, probably 40VA.

[rankot]

After putting 220/12V 40VA toroidal PTF as a second stage, I can confirm that B+ stabilise at 180V under load (measured just before preamp). It's probably OK. Also, I can confirm that everything is working right now, I put Celestion V-Type 8 Ohm speaker and it sounds OK. However, it is distorting bass frequencies, so probably something is farting. I will try with different preamp.

[anotherjim]

...resistance of a coil is approx 20 ohm. Is it OK - speaker is declared as 8 Ohm?...

I've never known a coil resistance to read that much higher. It points to something wrong. If a voice coil itself burns out, it usually fails completely open circuit.

Behind the coil connection tags, there are braided "flexi" conductors that attach through the cone to go towards the voice coil at the centre. The flexis often have a non-conductive core running inside the copper braid that provides mechanical support. Eventually, the braid frays and breaks down, either at the soldered end or where glued to the cone. That means the electrical connection can fail while the flexi appears to be intact. I've managed to repair some speakers that failed like that, but its seldom pretty and I wouldn't want to gig with that situation.

[thermionix]

A buddy's speaker quit on him, he had me check it out.  DC resistance reading was variable, but usually around 28 ohms or so (8 ohm speaker).  I think it was a failing internal connection (I was reading from the tinsel wires), rather than a blown voice coil.

Another possibilty is a blown (open) voice coil that you can still read a resistance across because the charred spot on the former is basically carbon, somewhat conductive.

Weird shit happens.

[rankot]

I tested the speaker on HiFi amp and it works there - works very good actually. So, maybe Jim is right - I will check those wires.

[rankot]

Jim, you were right - positive wire on my 15" speaker was damaged and almost broken! I've fixed it and now it is working fine.

Now I can get back to amp troubleshooting.

[rankot]

Going through the 470 ohm resistor is no big deal, just gives better filtering. The datasheet says that at 150V the current to the plate can be 18 mA, the screen at 3mA, triode say 2mA for 23mA. Ohm's Law says the voltage across the 470R resistor is 470 x 0.023 = 10.8V. So either you are drawing a tone of current to get a 100V drop or you have the wrong resistor value in there. The datasheet says that the grid of the pentode should be biased at 2.1V, Ohm's law again, 2.1 / 120R = 17.5mA. So that resistor value is appropriate. It would be good if you take dc voltages on the tube pins and for good measure across the resistors.

Just as a guess, you might have the input of the pentode shorted and it is drawing maximum current, heating up the OT. We need those voltages to eliminate a number of possible faults.

I'm running this at 220V, not 150. And I put 470 ohm to provide more filtering on B+ and to drop it a little, because my PS gives more than needed. I had to use 5k OTF, so I chosen this B+ to avoid excess anode dissipation.

Datasheet gives those typical values for PCL86:

datasheet	my schematic
Va 230V	220V
Vg2 230V	213V
Vg1 -5.7V	-5.2V
Ia 39mA	37mA/td]
Ig2 6.5mA	6.7mA

So it seems that my values are very close to defaults? However, the sound is distorted even at the very low level (especially bass), so I must try it with another preamp.

[anotherjim]

Have you looked at the triode section? Anode & cathode voltages? Maybe it's too close to cut off, bass having stringer amplitude, it clips first.

[rankot]

I'll check triode section too, but I am too busy those days with my regular job.

Maybe heater is too cold? PCL86 requires 13V/300mA, I have ~14.7V available from my first TF, so I used 6.8 Ohm resistor to drop it to 12.6V. I presume this filament voltage is within tolerances?

Regarding B+ filtering, this schematic also use 470 Ohm resistor BEFORE OTF and everything else, so I presume I wasn't wrong with that.
http://danyk.cz/elzes_en.html

[thermionix]

Regarding B+ filtering, this schematic also use 470 Ohm resistor BEFORE OTF and everything else

Yes, but additional filtering for the preamp stages.  My point earlier was that your filtering was basically reverse of what is commonly done, more filtering for the power stage than the preamp.  You can get away with a resistor before the OT in low power Class A amps.  I understand wanting an extra R/C filter for a single-ended (non-humbucking) power stage.

As an aside, the early Vox AC15s had a filter choke before the OT, but they had to use an extra beefy one.  They didn't do that long.  It was one thing I didn't copy when I built myself one.

[anotherjim]

Some resistance out of the rectifier is handy for checking total current via a voltage drop calculation. I fit 100R so I can work out V/R in my head easier.

Yes, your R6 filter to the triode preamp is small. They usually are some K ohms. If it drops too much voltage for your off-scheme pre-amp stage, I'd feed that from before R6 with its own RC filter.
Modern designs have relatively cheap access to >100uF caps compared to the 20-50uF in the traditional amps, so you can choose better filtering or less volt drop. I'd think it's whatever suits the design.

Whatever you do, some hum might be inevitable.

It might be worth running the heaters a little hotter, if only to rule that out.

[amptramp]

I could see duplicating the 10K / 22µF decoupling for the triodes so the low-level stages had more decoupling and also less coupling between the left and right sections.  I have seen amplifiers where separate rectifiers and filtering are used for left and right and the crosstalk is reduced even further.  This only matters if the maximum resistance to crosstalk is required.