Testing Tubes

[gbono]

I just bought a Sunn Sonaro chassis off CL and I thought I would test all the tubes before a bring it up on a Variac. The chassis looks "unmodified" and the previous owner had the PS filter cap (section) replaced 15 years ago. The amp hasn't been used since 2000.

Local electronics store has a tube tester - should I trust the tube tester or just ramp up the mains voltage while I check voltages??

[Enzo]

Tube testers can tell you a tube is definitely bad. They cannot tell you a tube is definitely good.

I fix tube amps every day. I don't admit it to customers, but I DO have a tube tester. I can't recall the last time I dug it out.

Power up an amp without tubes. Meter the socket pins. Are the voltages more or less what you'd expect? Try it with tubes then. If they freak out, try new tubes.

[gbono]

Enzo,

I have attached the schematic for the amplifier. What are the correct voltages on the EL34 and GZ34 supposed to read? THX

[Enzo]

Says right on there you should have about 455vDC at pins 2,8 of the GZ. Your voltages will be higher with no tubes installed, so if it is 500, so be it. The EL34s draw their power through the output transformer from power node A, 445v. SO I'd expect to find roughly that (higher without tubes) on pins 3 of the power tube sockets.

I'd expect similar voltage on pin 4 of each. In fact when teh tubes ARE installed, I'd probably expect pins 4 to be a volt or two higher than pin 3. But all we care about at this point is ballpark.

Then the bias is on pins 5 of the tubes. Schematic say -34v. SO see if you have more or less that at pins 5. WHile you are there , twiddle the bias adjust control and see if the voltage adjusts higher and lower.

If all that is in the ballpark, I'd try it with tubes. if any tube lights up real bright inside, turn it off.

[gbono]

Enzo, Thanks for the help. I assume that pins 2/7 are the heater for a EL43 - should I see 6V there?

How do I set the bias on the output tubes? I assume that I'm adjusting the bias pot while reading the voltage at the node between the two 270 ohm resistors??

[Amp Kat]

No ,those are the bias feed resistors and will not give you the tube current. If I were you unless you have a bias rite I would break the Cathode connection from ground on a power tube and connect a 1 ohm 1 watt flame proof resistor from cathode to ground. Measuring across this resistor in milivolts will give you the current thru the tube. I'd be looking for around 35 mv's and there is a 10k bias pot that sets it somewhere in there. You should be able to find it quite easy. 1mv will equal 1 ma of tube current.

[gbono]

I bought a bias tester from Amp-head. I looked at a data sheet for the EL34 and I would like to know what the quiescent current should be when setting the bias - 35mA looks pretty low?

My work background has been semiconductors. One vey crude observation is that the transfer fuction of a EL34 pentode "looks" alot like a JFET (VERY high voltage depletion mode). If there is no grid voltage applied the plate current is at a maximum = IDSS on a JFET.

[Enzo]

Parallels like that are useful for understanding signal flow and such, but trying to infer circuit details like that is going to get you confused. Yes, in the roughest sense it works like a JFET.

Idle bias curent is calculated by the voltage across the tube - plate to cathode - times the current through it. That yields power dissipated in watts. Many guys follow the 70% "rule." Set the current for whatever puts the tube dissipating about 70% of its maximum rating. SO if your EL34s have a max rating of 25 watts, then 70% of that is 17.5 watts.

Please don't worry about the half watt, bias is far less critical than so many people think. Your amp will work fine at 60% or even 50% dissipation at idle. On the other hand, some guys report the best tone they ever got was when their tubes were red hot and melting - sounded great for about 10 minutes.

I can;t tell you what current to set it for, I don't know your B+ voltage.

[Amp Kat]

500 volts x .035 = 17.5 watts. If your B+ is higher than that than yes it may be a tad cold but I would think that amp is right there. After looking at that schemo it's showing 455 and with todays voltages I'm thinking around 470 to 485 which would put it at (let say 475) 17.5 / 475= .036 ma's. 17.5 / 455 would be .038 so it's right around there give or take a few ma's.

[bnwitt]

Well I like to do as Enzo says. Start up with no tubes first. Check all voltages. Then start up with the tube rectifier if it has one and check voltages again. With a solid state rectifier the second step is not neccessary. I will at this point set the negative bias voltage to full negative before putting in any power tubes and make sure I have it present at each power tube socket's pin 5.

Then I will start up an amp with known good tubes from my stock on a light bulb current limiter to make sure there aren't any shorts. After that, start up without the limiter and start replacing my tubes with the original tubes. If an original tube causes a failure of some type its bad. It wouldn't hurt to go test your used tubes and eliminate any for sure bad ones before hand.

I have attached a spreadsheet for EL34 biasing. You must check the plate voltage and milliamp draw after each adjustment. as they will both change. Dial in a combination that matches a 70% reading on the chart. Be careful with your meter probe and don't short out any pins. Make sure you have a speaker plugged in or a dummy load whenever you turn on the amp with power tubes in it.

[gbono]

Excellent information - thanks!

The output tubes are Sylvania 6CA7/EL34 but one of the tubes has a broken alignment post - as does the rectifier tube. I purchased new output tubes from an online supplier AmpHead(?) as a spare set. There appear to be multiple on-line sources for tubes - is one source any better than the other?