Mystery Bias voltage on Deluxe Reverb 74

[Shannow985]

Hello everybody,

I'm new here and hope somebody can help me.
I just acquired a Deluxe Reverb from 1974, an import version (France here) and wanted to swap the aftermarket power transformer to a more conventional one from TAD.
Everything looks straightforward except the input bias voltage. From what I understand the PT should feed 50v AC to the 470ohm resistor before the diode (blue/red lead). The PT that is currently installed doesn't have a 50v output and the 330v output was used instead (Y lead split with the main rectifier section PT output : 308v was measured).
In order to do this apparently a bigger resistor was installed to drop voltage 33k + 47k(<- where the 470ohms should be). So far so good.
Where I'm baffled is that I measured 257v AC where I should input the 50v (blue/red PT lead) and I have 197v AC before the Diode... After the diode I have -48.5v DC and -34.6v DC on bias pot which is correct, everything works fine ! A 22k has been put one the bias pot instead of the 10k as well I should mention.
Everything else on this PT is fine and should be straightforward to swap the new PT. The rest of the deluxe reverb is standard for it's era (A1172) apart that a SS rectifier was used instead of Tube but that's another story.
I fear that putting the new PT 50v where at the moment I have 197v will upset thing (putting the 470ohm back to specs + 10k on the bias pot).

I do have pictures of the reading + circuit but they're dark as it's night time here ATM. I can take them again tomorrow in daylight.
Many thanks in advance

Jonathan

Edit: Ok it's called bias tap what i'm dealing with.

 

[Dacious]

Before the diode you're measuring AC. After the diode you're measuring it's been rectified down to a low negative voltage and it's a half wave rectifier which is very inefficient. You can't see it but much of that high input voltage at very low current is being lost at the diode as heat energy.

 

[Paul G.]

To use the 50V tap, you have to return your bias circuit to match the Fender schematic for your circuit, so 470 ohm dropping resistor and 10k to ground on the pot. There should be no problems at all. Just to be safe, check your bias voltage range with no power tubes installed. Fender shows -35V, so make sure you have a nice range that has -35 somewhere in the range. Before installing tubes and biasing, set the bias to the maximum negative voltage. Make sure all the knobs are set to zero and adjust the bias after the amp is nice and warm.

 

[Shannow985]

Thank you for both your help ! So everything is ok.
I can't figure how the technician that did this knew what resistor to use to be ok as the formulas I know don't really work here.
Current going through the diode must be quite different going from a 197v input bias tap to 50v input bias tap.

 

[String Theory]

Thank you for both your help ! So everything is ok.
I can't figure how the technician that did this knew what resistor to use to be ok as the formulas I know don't really work here.
Current going through the diode must be quite different going from a 197v input bias tap to 50v input bias tap.
If it's a diode specs related phenomenon should I replace it too ?

I'm going to offer the following advice as a starting point for searching more detailed info. If you have limited experience with this, you should be able to find good instructions if you have the right key phrases for searches. Hopefully a few guidelines will help.

The output tube -grids- won't normally draw much current, so the voltage divider calculations are pretty direct. Google 'voltage divider' to see how this works.

Take a look at schematics for regular Deluxe circuits to get an idea for resistor values. So you already have a pot for setting bias voltage? Good.

Output tube current can be monitored by installing 1 ohm resistors between each output tube cathode and ground (in place of the direct connection from cathode to ground). Measuring voltage across the 1 ohm resistor will tell you how much current each tube is drawing. The 1 ohm resistor will not interfere with operation of the circuit, and the voltage at the point will be very low (safe).

Do unplug the output tubes before testing the bias voltage. You don't want to blow a pair of tubes just cause the voltage is wrong. But also know that when you unplug the output tubes, the high voltages will remain on the B+ supply longer. As always, be careful.

You probably know this already, but more negative bias voltage will run tubes cooler. As the bias voltage gets closer to ground, the tubes will run hotter. Somewhere in there, you can find a spot where the tubes run reasonably warm, but don't overheat. Be sure that the bias voltage never gets accidentally grounded. Be sure that the circuit for the bias pot, will go more negative if the pot fails or gets dirty. You don't want the bias voltage going to ground under any circumstances. Tubes are tougher to get now.

Even if you think you've found a good bias point, turn off the lights and check the output tube plates to make sure there are no red/orange areas. I'm not talking about reflections from filaments. The plate itself will develop glowing areas if they're biased too hot.

If the previous tech used a rectifier (diode) that was spec'd for the higher voltages, the same diode should be fine. The voltage drop across them is pretty constant (around 0.65 volts). If you want to replace it, you can get a 1n4007 or similar for cheap. You don't need to worry about having a rectifier that has an over-spec'd "PIV" (peak inverse voltage), and a in4007 is about the same price as a 1n4001.

Let me know if you have more questions about any of my comments above.

 

[String Theory]

Thank you for both your help ! So everything is ok.
I can't figure how the technician that did this knew what resistor to use to be ok as the formulas I know don't really work here.
Current going through the diode must be quite different going from a 197v input bias tap to 50v input bias tap.

PS: There will be very little current going through the diode in that circuit. In many cases, the key spec for a rectifier will be "PIV" (peak inverse voltage), which has to be spec'd higher than max voltage at that point. But you're going from higher to lower, so the rectifier is probably fine.

Also be careful of polarity of the filter caps in the bias circuit. Positive side gets connected to ground, since the bias voltage is negative.

 

[Silverface]

I'd like to back up the "another story" and the SS rectifier, which can result in all sorts of odd, unexpected results.

Simple questions: 1. Why and 2. What are the exact numbers it's pumping into the appropriate amp sections?

Then I'll jump ahead - after getting the very rough/but safe negative bias voltage WITHOUT tubes - how are you planning on setting the output tube bias? (praying a calculator or oscilloscope isn't the primary...or only...tool...)

 

[Shannow985]

Many thanks ! at the moment I lowered the bias for the tubes. Apparently they're not matched as one what pulling 32ma and the other 26-28ish (plate voltage I think was 385v). Tried my Princeton's one which are matched and lowered them to 22ma.
I think I need to show pictures of what I have which will be worth 1000 words.
I will have an oscilloscope by tuesday

 

[Shannow985]

At the bottom left on the last picture, this is a 100uf cap

 

[Silverface]

Tried my Princeton's one which are matched and lowered them to 22ma.
I think I need to show pictures of what I have which will be worth 1000 words.
I will have an oscilloscope by tuesday

You DO understand amp safety and how to discharge high-voltage caps, right?

Get rid of the bias cap first - it's about 25 years too old! While you're at it, get rid of the carbon comp resistors in the filament circuit and install matched metal film resistors.

Why did you set the bias to 22ma? Did it sound best? If so and the tubes aren't red plating - which is very unlikely if the plate voltage is around 22ma - it's probably fine. But if you know the plate voltage you can set the bias - using standard tube specs - for around 70% dissipation, then play/listen and make sure you're not red plating...then tweak a few ma up and down and find the sweet spot. If it's not red plating at that point you're done.

You only need a scope to trace the circuit for problems - not to set the bias. The crossover-notch method doesn't factor in sound. Neither does JUST setting by dissipation percentage.

You have to LISTEN to the amp being played and normal volume.

And you MUST know the plate voltage. You can't set the bias in a safe range for the tubes without knowing the plate voltage.

What is the VOLTAGE rating of the 100ufcap, and how are all the filter caps wired in the circuit? And how old are they? If over 15 years - change them! 15-20 years is the maximum service life of filter and bias caps.

 

[Shannow985]

I have bought a few 5w 470ohms resistors to use them to discharge the caps using one as a probe to avoid sparks (isolated wire from ground) and then double check with multi-meter.

I just checked my previous comment on the plate voltage I confirm this is 385v plate voltage (probe on pin 3). At 22ma this gives me 60% dissipation on a 14w max I believe (0.022x385=8.4w). It was to be really on safe side with the good 6v6gt of my PR's because current 6v6gt DR's are clearly unmatched and I have 20ma on one and 24ma (66% dissipation on that one and that the one probed for voltage). It did sound good on 30-32ma though. Just checked at this setting the plate voltage is 376v now so we have 86% plate dissipation isn't it too much ? (0.032*376=12w and the other tube is at 24ma)

Don't know what this cap on the bias pot does and was going to remove it to put the amp back to spec as it's not on the schematics, thanks.

The carbon comp resistors on the filament circuit you are referring to are the two 100ohms near the pilot light ? (was going going to replace them as they looked tired)

the 100uf cap is rated at TS450v vcc and TP550v vcc.

22uf I can only read the left one.

 

[String Theory]

> Don't know what this cap on the bias pot does and was going to remove it to put the amp back to spec as it's
> not on the schematics, thanks.

Not sure which schematic you're looking at, but you need filtering on the bias supply. Were there two separate caps?

Make sure that filter caps in the bias supply are good quality. If they fail, your bias supply could be grounded. I had already pointed out that would not be good, so don't cut corners there.

 

[corliss1]

I don't like the looks of anything under that capacitor can. I'd replace all those just out of principle.

 

[String Theory]

The crossover-notch method doesn't factor in sound.

Agreed. I still see this recommended once in a while and it doesn't make sense. Often the tubes get biased very cold. And of course tubes can drift in either direction. Best to bias in a safe zone, and if there's some perceived crossover distortion (doubtful) you'd need to live with it anyway; otherwise dissipation could be exceeded. (I believe we agree on that)

I've even heard some crazy recommendations to turn bias until the tubes red-plate, then back it down until the red spots disappear. Totally irresponsible.

 

[Dacious]

With your two apparently mismatched tubes, try swapping them between sockets. Often, due to different resistance in the two halves of the output transformer you will find tubes that are 3-4ma apart will close up on current draw if swapped. If the disparity gets worse, then you may consider them mismatched.

If you measure in six months you're likely to find bias on each tube has shifted.

 

[Shannow985]

> Don't know what this cap on the bias pot does and was going to remove it to put the amp back to spec as it's
> not on the schematics, thanks.

Not sure which schematic you're looking at, but you need filtering on the bias supply. Were there two separate caps?

Make sure that filter caps in the bias supply are good quality. If they fail, your bias supply could be grounded. I had already pointed out that would not be good, so don't cut corners there.

Ah ! I was talking about that one (47uf 100v):

I'll replace the bias one as well then but I can only find 25uf at 50v at the moment which is a bit low voltage from what I read (Jupiter Cosmos Wet Electrolytic type is ok ?).

I'll try the trick of swapping the two power tube for bias thanks !

I checked the whole board yesterday evening and everything is spot on the A1172 schematics except the two missing capacitors for the cathode bypass (v3A and V5A) and a resistor at the same place.
Apparently it's common on Silverfaces of that era (seen a few inside pictures of 1974-75ish without them on google+ resistor value identical).

 

[Shannow985]

Is this a zener diode ?

Does this explain why 197v input gives me 50v output ? (it's used as a regulator?)
Can I use it with the classic PT 50v bias tap ? Can't find the specs it's written RO (RC?)3060...
It looks like this is original should I keep it or replace with 1n4007 to be safe?

 

[corliss1]

Since you're replacing the PT and bias network, just replace all that so you're sure.

 

[Silverface]

I've even heard some crazy recommendations to turn bias until the tubes red-plate, then back it down until the red spots disappear. Totally irresponsible.

That's almost a direct quote from Gerald Weber's first book. " A hip Guide bla bla"...

IGNORE it!!

(I somehow fouled up his quote, but basically is two tubes are in the safe bias range but a few ma mismatched, it's no big deal. I usually use 5ma "off" as my mismatch point, but it depends on the tubes, amp, plate voltage etc. Many amps sound better withy the tubes slightly mismatched - and even if matched You'd have to check both sides of the OT and both sides of the phase inverter for perfect balance as well.

See the non-quoted quote above!

I'll replace the bias one as well then but I can only find 25uf at 50v at the moment

DON'T replace it with that! it can't handle the voltage. leave the old one in plqce and order a 100uf/100v cap (a 50uf/50v or 50uf 100v will all work the same - but a 25 V cap can't handle the negative bias voltage. The 100/100 is what many techs use as a standard bias cap.can handle current and voltage - you can always go too high on bias caps and it affects nothing - except for giving you a higher margin of safety.

When you replace it pay attention to the polarity!!!

Is this a zener diode

No. It's a standard diode.

Many of your questions are SO basic 1)I'm a little concerned about you understanding amp safety and high voltage caps - in some amps they can burn or KILL you with the amp turned off, unplugged and stored for a month!

And not recognizing a basic, common diode used in almost every Fender fixed bias amp of that era is worrisome - I very seriously suggest you take the amp to a qualified tech and learn basic electronics, parts identification and work on simpler things until you build your knowledge.

You can really (expensively!!!) screwup an amp touching one wrong part or connecting a polarized part backwards - or seriously injure yourself.

 

[Silverface]

22uf I can only read the left one.

I just saw the filter capacitors.

Even ifn they had date codes they should be replaced. One of the most precarious soldering jobs I've seen in over 30 years If you look at the capacitor on the left, the top wire runs up, curves to the right and is simply laying on ntop of another wire. There is ONE tiny cold soldering joint used as glue - and barely holding them together.

Wire is not glue. Itn is used to secure already-solid mechanical connections. There are eyelets that the installer simply ignored - the wires SHOULD be inserted into the eyelets- the mechanical joint - and THEN soldered. The solder should be bright and shiny - sign of a good joint. Dull solder is sign of a cold joint - but there's not much of a viable connection anyway.

This amp has so many issues, modifications, changed parts etc. It is NOT a good project for a DIY'r with little experience. There are too many thing that could go wrong - and iy's impossible to tell you where to start, because there is no one starting place.

It needs to be taken to a professional amp tech - it will undoubtedly cost you LESS money in the long run and you'll end up with a working amp; not one with repeated problems.