AB763 Deluxe Reverb

[NickC]

Howdy,

Working on a home-brewed Fender Deluxe Reverb AB763 (schematic attached). I blew a 1 slo-blo amp fuse. Replaced with a 2 amp slo blo (after doing some inter web research). Then the 6V6GT output tubes red plated. I realized quick and shut down power. Took out 2 amp fuse and put in new 1 amp slo blo fuse.

I think the voltages all seem too high. Before the fuse popped, the amp was working (all functions), but making an intermittent noise (noticeable feedback like hum).

Here are the specs:

Mercury Magnetics Blackface Deluxe Reverb FDP20 PT (200 mA)

GZ34 Rectifier putting out 483 VDC

B+ String (as per schematic):

(A) 460 VDC
(B) 458 VDC
(C) 366 VDC
(D) 307 VDC

Bias (max negative to prevent red plating) -60 VDC

V1 7025 = preamp normal channel
pin 1 plate 200 VDC
pin 2 grid 0 VDC
pin 3 cathode 0 VDC

pin 6 plate 209 VDC
pin 7 grid 0 VDC
pin 8 cathode 0 VDC

V2 7025 = vibrato (trem) channel preamp
pin 1 plate 214 VDC
pin 2 grid 0 VDC
pin 3 cathode 0 VDC

pin 6 plate 208 VDC
pin 7 grid 0 VDC
pin 8 cathode 0 VDC

V3 12AT7 Reverb Send
pin 1 plate 454 VDC
pin 2 grid 0 VDC
pin 3 cathode 7 VDC

pin 6 plate 453 VDC
pin 7 grid 0 VDC
pin 8 cathode 7 VDC

V4 7025 half reverb recovery, half vibrato channel gain
pin 1 plate 213 VDC
pin 2 grid 0 VDC
pin 3 cathode 0 VDC

pin 6 plate 199 VDC
pin 7 grid 0 VDC
pin 8 cathode 0 VDC

V5 Vibrato (trem) circuit
pin 1 plate 447 VDC
pin 2 grid -45 VDC
pin 3 cathode 0 VDC

pin 6 plate 401 VDC
pin 7 grid -44 VDC
pin 8 cathode 0 VDC

V6 12AT7 Phase Inverter
pin 1 plate 213 VDC
pin 2 grid 47 VDC
pin 3 cathode 75 VDC

pin 6 plate 209 VDC
pin 7 grid 50 VDC
pin 8 cathode 75 VDC


V7 6V6GT (Electro Harmonix)
pin 3 Plate 463 VDC
pin 4 Screen Grid 460 VDC
pin 5 Control Grid -59 VDC
pin 8 Cathode 0 VDC

V8 6V6GT (Electro Harmonix)
pin 3 Plate 464 VDC
pin 4 Screen Grid 462 VDC
pin 5 Control Grid -59 VDC
pin 8 Cathode 0 VDC


So, what do you think ...... bad tube? Are the voltages too high. Should I adjust the dropping string (maybe replace one of the 10K R with 18K R)?

I greatly appreciate any insights. Thanks!

[martin manning]

460V is pretty hot for 6V6's. Did you measure idle current to get plate dissipation? What is the OT primary impedance? Got a 5Y3? ;^)

[NickC]

460V is pretty hot for 6V6's. Did you measure idle current to get plate dissipation? What is the OT primary impedance? Got a 5Y3? ;^)

I didn't adjust the bias yet, just dialed the negative bias voltage all the way up to -60 VDC. Figured I'd probably have to loose about 50 volts or so to the PA and drop the preamp volts by 20 or 30 to get close to the published specs on the Fender layout.

Output Transformer is Heyboer HY041318 (HTS-8962) Deluxe Reverb replacement. I believe it's a 6.6K @ 8 ohms.

I don't have a 5Y3 or a 5U4GB. I do have a pair of Tung Sol 6V6's on the way (don't trust the ones in there now, since they red plated).

I'll have to dig through my parts stash to see what I have to adjust the dropping string.

[martin manning]

5U4 has a 3A filament, where 5Y3 is 2A like GZ34. At that voltage and 6k6 load you are in danger of over dissipation. JJ 6V6S might stand for it, but it looks scary.

[sluckey]

V1, V2, V4, and V5 ALL have zero volts on the cathodes. That means none of those tubes are passing any current. Look for a common problem, such as no ground connection for those tubes. Measure resistance from those cathodes to ground. One probe directly on the socket pin, the other on chassis. Probably will show open circuit at this point.

[NickC]

V1, V2, V4, and V5 ALL have zero volts on the cathodes. That means none of those tubes are passing any current. Look for a common problem, such as no ground connection for those tubes. Measure resistance from those cathodes to ground. One probe directly on the socket pin, the other on chassis. Probably will show open circuit at this point.

Initial voltage measurements were with vol pots all down, no signal at input, and 8 ohm dummy load on spkr jack.

I traced continuity from each of those pin 8 cathode connections to their respective cathode caps/resistors, and from the other side of those to ground. Grounds are intact. Checked the resistors in circuit with Fluke and they test to spec. I'd have to lift the caps to test them (all Sprauge's).

Measuring for resistance from the pin 8 to the cathode caps/r's, and from the other side of those cathode caps/resistors to ground yields very small resistance ranging from 0.3 to 0.5 ohms.

Thanks for guiding me through that test. I'm a novice and have much to learn.

[martin manning]

The high voltage on the power tube plates is due to the low current draw that you are experiencing right now- the output tubes essentially shut off with bias voltage at -60.

I'm confused by your cathode voltage readings on V1, V2, and V4, though. Power supply node D is at 307, and all of the plate voltages are around 100V lower, which says that there is ~1mA flowing through each of those triodes.

V5's plate voltages are close to node B. They should be at this point because both triodes in that tube are in cutoff because of the negative bias voltage being applied to the grids when the tremolo is off.

Don't worry about tweaking voltages yet. The first thing I'd do is find a way to set the bias on the output tubes correctly. Did you install 1 ohm current sensing resistors from the 6V6 cathodes to ground?

[Stevem]

Your red plating of the outputs with -59 on pin 5 is whats nagging me?
If the outputs only red plated for a few seconds they are likely still good enough to test the amp with, but the first thing I would do is confirm that they are not shorted.
A ohm meter test from pins 2 or 7 to pin 5 and pin 3 and pin 4 should not show any resistance, if they do than that is the cause of the blown fuses.

[NickC]

I didn't install 1 ohm bias sensing resistors in this build.

The bias was initially set using Weber Bias tool and their calculation app on website, biased at a bit under 70% dissipation (a little cold). It was at that setting when the 1A fuse popped. When I replaced with a 2A slo-blo fuse both 6V6's red-plated. Replaced the 2A fuse with a good 1A slo-blo (metered for continuity).

When I pulled the chassis I increased the negative bias voltage to max. Then removed all tubes and checked for shorts, and measured pin voltages, B+ string, rectifier output, etc. Next put the preamp, rev, vib, PI tubes back in and checked voltages again. Finally, I put the 6V6's back in and measured again.

I'll re-bias and see where it sits then. Weber Bias Calc shows:

Class AB 6V6GT Tubes Plate 460 VDC -----> bias should be at 21.3mA

5Y3 rectifier is sounding like a good idea. I'd like to lower the voltages overall. I first have to check whether the PT can supply sufficient current for the 5Y3. (Update: ordered 5U4GB and 5Y3 rectifier tubes)

Thanks for the help ya'll!

[martin manning]

If you do put a 5Y3 in, it will have the same filament current as the 5AR4/GZ34, so no worries there. A 5Y3 is only capable of about half of the current that a 5AR4 can handle, though.

You measured voltages at the power supply nodes with all of the tubes removed- that will tell you if you have some current draw that shouldn't be there. You should find them all close to the same value. If that checks out then put just the power tubes in, set the bias to around 70%, see if it remains stable and where the plate voltage ends up.

[NickC]

If you do put a 5Y3 in, it will have the same filament current as the 5AR4/GZ34, so no worries there. A 5Y3 is only capable of about half of the current that a 5AR4 can handle, though.

You measured voltages at the power supply nodes with all of the tubes removed- that will tell you if you have some current draw that shouldn't be there. You should find them all close to the same value. If that checks out then put just the power tubes in, set the bias to around 70%, see if it remains stable and where the plate voltage ends up.

Thanks Martin for the 5Y3 tip!

The 5Y3 brought the voltages down to a more comfortable level. 430 VDC now on the 6V6 plates. Biased at 19mV on one, and 18.8mV on the other (a tad on the cold side .... Weber Calc showed 19.3mV with plates at 430 VDC).

The preamp, trem, rev and PI tubes are now closer to original Fender layout specs (a few volts high here and there).

Put it all back together and play tested it for a while with a hum bucker LP straight in, keeping an eye on the tubes. Everything works. No red-plating. Turned it up to 6 on the Trem/Rev channel and it speaks with authority. Even turned up that high it's quiet, with no obnoxious humming or buzzing .... can barely hear anything from the amp with the strings damped. I'll keep in in the rehearsal room for a few weeks and let it burn in good, and keep an eye on it to make sure it's behaving. But I think it's fixed.

Thanks Martin, SteveM, and slucky for the guidance. Much appreciated!

[NickC]

By the way ..... now that the DR is working I'm feel hungry for starting another build. Is that normal?

[Colossal]

By the way ..... now that the DR is working I'm feel hungry for starting another build. Is that normal?

Completely

[martin manning]

The 5Y3 brought the voltages down to a more omfortable level. 430 VDC now on the 6V6 plates.

Going from 460 to 430V sounds like maybe you have a Russian 5Y3 in there now? If that is the case, a NOS or a new JJ will take the plate voltage down another 20 volts or so.

[NickC]

The 5Y3 brought the voltages down to a more omfortable level. 430 VDC now on the 6V6 plates.

Going from 460 to 430V sounds like maybe you have a Russian 5Y3 in there now? If that is the case, a NOS or a new JJ will take the plate voltage down another 20 volts or so.

The one I got is a new production JJ 5Y3S.

I looked on eBay at 5Y3 and there is a dizzying array of 5Y3 and 5Y3GT tubes with prices all over the map (Phillips, GE, Sylvania, RCA, National). Also came across references to the Bendix 6106 (5Y3WGT), which is much more expensive. Any of those particularly good?

Thanks Martin, I appreciate your insight and advice!