princeton reverb II

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ward

Well-known member
Joined
Jun 4, 2004
Messages
348
Location
Belgium
Hi,

I am looking at a fender princeton reverb II, it needed a retube and I redid all the electrolytics caps.
But there must be something else wrong, something in the bias arrangement keeps overheating.
http://www.stratopastor.org.uk/strato/amps/prii/schematic/schematic.html

Could somebody explain the 2x 390ohm resistors in the bias ?
I replaced it with two parallel 1K8 1K5 1W resistors.
But they overheat with all the tubes out.

There's a drop of 50V over 780 ohm, thats 0.064A, so 3,2Watt ?
Is this underdesigned, or am I getting it wrong?
 
Last edited:
Schematic says two 390 ohm / 2 watt in series, which is 4 Watts.
Your replacement is undersized - order the right parts.
These current limit into the 6.2v zener diode
 
Thanks for the reply.
I went for 5x 3k9 but it’s still getting way hotter than I like, almost too hot to touch.
The voltage drop I measure is 55v.
55x55/780=almost 4W
Maybe I better go with 10W.
Or I probably could go a little higher than 780 ohm.
 
Five 3k9 in parallel is 780, and if they are all 1 watt the total power capacity is 5 watts. But they need to be spaced to have full convective dissipation each. Not touching each other. The power rating of a resistor is based on the outer surface area and temp capability of the material. Modern power resistors are smaller because they are intended to run hotter. Nothing wrong with going to 10 watts if you have the space
 
the wire wound sand blocks can run hot enough to boil water, but it is the PC Board traces that have a hard time with the heat and vibration.
I usually mount those two resistors on solder lugs off the board and run wires to the board, and you can up the R to 470 with no voltage dropout, even higher if you want, 470 is a common value for sand blocks, you can get bigger zeners also, but probably stock will be ok, but that part of the circuit has nothing to do with pwr tube bias, are you having problems there too?

you can use 10W if you go off board>

r.jpg

this secti
 
Those resistors often come with a metal bracket that bolts directly to chassis , suspending the weight of the actual component by its legs might make it prone to vibrations , I have on a few ocassions found sand box style resistors where the sand/cement had crumbled and the ceramic box had fallen away .
 

Attachments

  • resistor.JPG
    resistor.JPG
    10.9 KB · Views: 6
but that part of the circuit has nothing to do with pwr tube bias, are you having problems there too?
It had problems with powertubes redplating.
The old two 390 resistors measure way too low and seemed to have been hot. And I believe this heat damaged the tagboard, which fell apart when I started desoldering. Then I noticed that one side of the biaswinding had come loose.
The 47ohm in the bias was burned.
I replaced that one and the diode.
And the 6.2 zenerdiode.
I believed all those problems were linked and caused by heat.
 
Finally I had time to dig into this some more ..
I put in a 10W sand block as suggested.
Then I hooked it up to an AC regulator.

Voltages with only the preamp tubes in seem fine to me :

ACin 24vac
heaters 0,65vac
PTransfo out 31vac
B+ 66Vdc
A 65Vdc
B 64Vdc
C 64Vdc
Bias -5,5Vdc

ACin 120vac
heaters 3,3vac
PTransfo out 158vac
B+ 218Vdc
A 218Vdc
B 214Vdc
C 208Vdc
Bias -17Vdc

Then I put some power tubes in :

ACin 120vac
PTransfo out 85vac ?????
B+ 88vdc

Something is wrong, the powertransfo feels a little hot.
the output transfo, I can touch it for ten seconds before it becomes too hot.
I'm suspecting the output transformer has overheated and is bound for replacement?
Maybe the power transformer too ?

Are there some tests I could do to be more sure ?
 
Finally I had time to dig into this some more ..
I put in a 10W sand block as suggested.
Then I hooked it up to an AC regulator.

Voltages with only the preamp tubes in seem fine to me :

Are there some tests I could do to be more sure ?
Since the bias voltage is correct with no tubes, I would look for a shorted tube or a mistake when rewiring the thing.
The power transformer seems right to me (from a distance).
If the output xfmr was shorted it would not heat up.
 
If I understood correctly, you tested the amplifier at a maximum of half the mains voltage with and without output vaccum tubes. Such a test cannot be considered correct, simply because vacuum tubes are heated to only half the heating voltage that way, and are not defined for these operating conditions. Maybe your amplifier oscillates in those conditions. Have you perhaps tried to measure these voltages with an oscilloscope?
I would advise you to temporarily remove the -bias resistor used as a dropdown resistor for the drive opto resistor (2*390ohm 10W) and insert it in series with the anode choke. Power it up and monitor the anode current and -bias all the time. For safety, you can also short a 10k resistor (part of voltage divider with 15k) for -bias so that for testing purposes the output vacuum tubes will certainly be safe in class B as this will increase -bias. Before that, make sure that you have wired all the new electrolytic capacitors correctly.
 
So far I can't find mistakes in the components i replaced.
Electrolytics are wired correctly.
Output tubes work fine in another amplifier.

I would advise you to temporarily remove the -bias resistor used as a dropdown resistor for the drive opto resistor (2*390ohm 10W) and insert it in series with the anode choke.
Can I install 1 ohm cathode resistors instead? seems easier and safer.


edit, I just tried it on full mains voltage, couldn't help myself, and its oscillating on a low frequency :rolleyes:

edit2, I went looking for shorts, with the power tubes in its only 1.8ohm from pin 3 to ground ?
 
Last edited:
Since the bias voltage is correct with no tubes, I would look for a shorted tube or a mistake when rewiring the thing.
The power transformer seems right to me (from a distance).
If the output xfmr was shorted it would not heat up.
You were right, there's something with the power tubes that makes a short.
I used JJ 6V6S as replacements, because that's the ones I have.
When I put in the old original Fender 6V6GT's everything behaves as expected.

ACin 266vac
heaters 7,29vac
PTransfo out 342vac
B+ 468Vdc
A 466Vdc
B 423Vdc
C 380Vdc
Bias -60Vdc

It's all a bit high, but the mains power seems exceptionally high too.

Unfortunately my attempts at buying new 6V6GT's have failed for now ....
 
Replace the two cpupling capacitors feeding the grids of the output tubes. Leaky caps will drive the control grid positive and ultimately cause runaway of the output tube current. This is especially evidenced by the runaway nature of your situation. I have diagnosed this successfully before.
 
Replace the two cpupling capacitors feeding the grids of the output tubes. Leaky caps will drive the control grid positive and ultimately cause runaway of the output tube current. This is especially evidenced by the runaway nature of your situation. I have diagnosed this successfully before.
If it was the case, the original tubes would not bias correctly either.

The comment "with the power tubes in its only 1.8ohm from pin 3 to ground" in post #11 leaves no doubt about the cause of the problem. Short between plate and cath.
 
You're right I missed that detail. So unusual to have a short in an output tube. Would have shown up in a tube tester, I think.
 
Back
Top