Deluxe Reverb Build - Bias way too hot, and other problems

[waspclothes]

Hey all,
I recently picked up a 70's silverface Deluxe Reverb that had all sorts of problems, so I decided to start from scratch and rebuild it with the classic blackface AB763 circuit, replacing all resistors, capacitors, etc. I'm pretty familiar with electronics, but this is my first attempt at building an amplifier.

After spending quite a few days wiring things up, checking and rechecking.. I recently powered it up. I was very happy to find out that the amp did work, including the reverb. The tremolo however didn't work, but more on that later. Immediately after powering up the amp, I noticed that the tubes got VERY hot and very fast, with the metal plates inside starting to glow bright red in 30 seconds.

Today I tried biasing the amp and ran into some problems. To bias the amp, I built my own version of the Weber bias rite head (http://taweber.powweb.com/store/br-kit-diagram.pdf) so I can easily measure the bias current (through a 1ohm resistor) and the plate voltage (between a 1M and 100ohm resistors) with my multimeter.

Plate voltage measured 305V (30.5mV between the resistors), but the bias current steadily climbed to over 130mA (130mV across the resistor)!

So I figured something wasn't right, so I went and measured a few voltages inside the amp. From the wiper of the bias pot, I was measuring +12mV but I'm supposed to be measuring -35V here apparently). The red/white wire from the transformer that goes to the bias network I measured at +29mV. B+ voltages are 319V and 326V.

If it helps, my bias network consists of a 470-1W flameproof resistor, the original silverface diode and a 47uF 160V capacitor.

Anyways, I have no clue what is wrong. Really hoping for some help here. I must have messed something up..

Maybe this is related, but the tremolo doesn't seem to work. I see the neon is on inside, although it doesn't appear to be oscillating (i'll check again) and turning up the intensity just turns the overall volume down.

I'll post some pictures tomorrow if that will help.

Mike

[sgelectric]

Is your bias cap oriented correctly? It should be + to ground.

[waspclothes]

Hey sgelectric, thanks for the reply
I've looked at the circuit several times checking the orientation and all things seem correct. I'll look at it with fresh eyes tomorrow and post a picture. I'll also pop the bias tap wire off and measure the output of that (expecting to see -50V).

[sgelectric]

Check the bias diode also

[MWJB]

What are your power tube voltages at pins 5. You can measure with the tubes removed to save redplating another pair.

We're not trying to set bias by pin 5, just ensuring that we are in the ball-park.

[Bruce / Mission Amps]

Pull the power tubes out and check the bias voltage again... at 130ma, YIKES... you might have a shorted power tube element... screen or plate to grid, etc.

also... you have to have the center pin-lug of the vibrato RCA jack on the chassis grounded for the vibrato work.
Use a skinny alligator clip in it to see if the oscillator turns on.

[tubeswell]

Everything in the bias supply needs to be reverse-biased. The bias supply diode has to have the cathode (banded end) pointing to the PT winding. The filter cap(s) in the bias circuit need to be +ve side to ground.

Basically you want to rectify only the -ve voltage swing from the AC supply and then stabilise it as a steady -ve DC voltage. For 6V6s in a DR, the (-ve) voltage on the grid pins should be around -35V to -38V DC*. Pull the output tubes out and fix it up until you have this voltage on Pin 5 of each 6V6 socket.

* i.e.: so that you get about 18-22mA tube current per tube (the 6V6s should ideally be idling at somewhere between 60% and 70% of maximum dissipation)

[waspclothes]

Hey all,
thanks for all the replies so far. I'm happy to say I found the problem (although the tremolo still isn't working, but that will have to be another post for another day).

Anyways, I ended up finding a glob of solder on the side of the bias pot that was grounding where the rectified bias comes in.

So I've corrected that and I'm now measuring:

385V plate voltage
42mA bias current.

I can't get the bias current any lower for some reason, so I'm guessing I change the resistor that's off the pot? Or do I change the resistor right after the transformer bias tap?

[tubeswell]

I can't get the bias current any lower for some reason,

Then you need to increase the negative voltage at the grid (in order to decrease the tube current).

so I'm guessing I change the resistor that's off the pot? Or do I change the resistor right after the transformer bias tap?

The resistor network in the bias circuit is a voltage divider, and it as a 'Voltage input', a 'Voltage output', and a ground return. You can increase the negative voltage at the Voltage output ('knee') of the voltage divider by either:

Increasing the tail ('lower leg') resistor on the ground side of the bias pot (which will shift the adjustable range 'upwards' i.e.: more negative); or

Increasing the resistance value (and hence the range) of the bias pot; or

Decreasing the other ('upper leg') resistor on the rectifier side of the bias pot.


The formula for voltage divider is:

'V out' = [lower leg/(upper leg + lower leg)] x 'V in'

where the lower leg is usually the resistor plus the bias pot, and the upper leg is the other resistor.

So say you are starting with 50VAC on the bias tap, which you have rectified and filtered to -69VDC (which becomes your 'V in') and you want -28VDC out. You need a voltage divider output that is 40.6% of the input voltage. You could use any reasonable appropriate combination of resistances to get this, provided that the resistors are not so small that they pass too much current to ground.

82k and 56k would work e.g.:

56k/(82k + 56k) = 40.579%

So would 100k and 68K

68k/(100k = 68k) = 40.476%

etc


For adjustable bias, you just split the the value of the lower leg. So say you were splitting the 56k - you could use either (say) 47k tail resistor + 20k pot, or 33k tail resistor + 50k pot, so that the pot and tail resistor gives you an adjustment range around the 56k. There you go

[MWJB]

Waspclothes "I can't get the bias current any lower for some reason, so I'm guessing I change the resistor that's off the pot? Or do I change the resistor right after the transformer bias tap?"

Increase thevalue of the resistor grounded from the bias pot. The resistor after the bias tap is only 470ohms, reducing this will make no appreciable difference. Select the resistor to give ~30mA at the hottest bias pot setting.