Mark III blue stripe redplating, quick diagnosis and repair

[woodbutcher65]

So I've been searching for a Mark III Blue Stripe, long head preferred, for some time....and it turns out that one of my friends who's an amp collector just happens to have one and it's not working right.

So we work out a trade agreement, subject to my being able to fix it.
He brings it over. I hook it up, plug it in, power it up, and the telltale hum of a massive power tube imbalance is evident in seconds. And I see both tubes on one side of the OT start to red plate.

There are FOUR possible reasons for red plating.

Lack of adequate bias current. With this amp, there's about a 0.0001 percent chance of this affecting only half the power tubes.
There's literally no component that can fail that would affect bias voltage to more than a single tube, those being the 220K resistors that
provide the bias to the outermost (EL34) tube sockets.

Shorted power tube. Not so likely that two will do it at the same time.
Open screen resistor(s) but they'd be on a per tube basis.
Shorted 0.1 uF coupling capacitor.

A quick side to side tube swap rules out bad power tubes but I'm not going to rely on tubes that I saw red plating.

Meter checks of the screen resistors indicates no problems with them.

The 0.1 caps are 400 volt rated Orange Drops. Overkill. WHY? That's only bias voltage in this circuit, 100 volt rated caps are more than sufficient.
I'm also suspicious because they ARE orange drops. Great capacitors, yes, but I've seen a lot fail when used as coupling capacitors to the power tubes.
So I replace the pair of 0.1s with a pair of Panasonic 0.1 250 volt caps, put the tubes back in, hook up, hit the power, and pray as only an electronics technician can pray.

Success. Bullseye. Amp runs great, just needs some pots cleaned. It's been running for a couple hours now and the trade has been done.

The point of this is to define red plating/bias issues, as pertaining to Mark IIIs.

There are only four causes of this problem. The amp is VERY easy to troubleshoot in any event there is redplating or a loud hum.

To reiterate:
Shorted output tube
No bias supply
Open screen resistor (or way high off value)
Leaky coupling capacitors.

I hope this saves someone a lot of time dealing with a cranky Mark III with a red plating tube problem.

 

[woodbutcher65]

Oh, just wanted to add, after pulling the pair of orange drop coupling capacitors and replacing them, I later measured them and one is a dead short, 0.12 ohms. The other still reads OK.

 

[user 41014]

If you're talking the two .1uF 400V coupling caps at the PI that is the 1st place to look for two tubes red-plating and IME the most common failure in old Marks. Continuity function of your meter will likely show a short in-circuit if anyone gets this red-plating blowing up the bias circuit for 1/2 of the power section.

630V:
Adding V rating always provides more component life but in this case it is a necessary change and Mesa upgraded the spec from 400V to 630V due to the failures here.

Get those 250V's outta there.

 

[Henz]

So I've been searching for a Mark III Blue Stripe, long head preferred, for some time....and it turns out that one of my friends who's an amp collector just happens to have one and it's not working right.

So we work out a trade agreement, subject to my being able to fix it.
He brings it over. I hook it up, plug it in, power it up, and the telltale hum of a massive power tube imbalance is evident in seconds. And I see both tubes on one side of the OT start to red plate.

There are FOUR possible reasons for red plating.

Lack of adequate bias current. With this amp, there's about a 0.0001 percent chance of this affecting only half the power tubes.
There's literally no component that can fail that would affect bias voltage to more than a single tube, those being the 220K resistors that
provide the bias to the outermost (EL34) tube sockets.

Shorted power tube. Not so likely that two will do it at the same time.
Open screen resistor(s) but they'd be on a per tube basis.
Shorted 0.1 uF coupling capacitor.

A quick side to side tube swap rules out bad power tubes but I'm not going to rely on tubes that I saw red plating.

Meter checks of the screen resistors indicates no problems with them.

The 0.1 caps are 400 volt rated Orange Drops. Overkill. WHY? That's only bias voltage in this circuit, 100 volt rated caps are more than sufficient.
I'm also suspicious because they ARE orange drops. Great capacitors, yes, but I've seen a lot fail when used as coupling capacitors to the power tubes.
So I replace the pair of 0.1s with a pair of Panasonic 0.1 250 volt caps, put the tubes back in, hook up, hit the power, and pray as only an electronics technician can pray.

Success. Bullseye. Amp runs great, just needs some pots cleaned. It's been running for a couple hours now and the trade has been done.

The point of this is to define red plating/bias issues, as pertaining to Mark IIIs.

There are only four causes of this problem. The amp is VERY easy to troubleshoot in any event there is redplating or a loud hum.

To reiterate:
Shorted output tube
No bias supply
Open screen resistor (or way high off value)
Leaky coupling capacitors.

I hope this saves someone a lot of time dealing with a cranky Mark III with a red plating tube problem.

Ty for the advice e WB!! In my DRG C+, I was measuring for any leaky Orange drops. When I got to the PI circuit, both plates measured app. 320vdc. On the other side of the .1's there was app. 60vdc. That surprised me till I realized there was bias voltage present from bias in. If you don't mind sir, can I give you some other voltage readings that I found in my PI and let me know if they are normal or if I have an issue. Both grids pins have app. 75vdc present, even after the first .1 Orange drop from input from eq (V5A). The input side of the first .1 from eq reads 0 mv. And 88vdc on cathodes (they are tied together). Are these voltages " back feeding" through the feedback loop? Blessings, Ty for your time!, Jim

 

[woodbutcher65]

Rather than go to all that trouble, just remove and test the orange drops out of circuit. They should not read as low value resistors.

I got a couple of these and they tell me the ESR and resistance and capacitance of any common capacitor. Get one or two.

https://www.ebay.com/itm/2754341298...Fol5tKVkq9V//orXKQc1zNbF2juD|tkp:BFBMgtiA19Jh

 

[Henz]

Rather than go to all that trouble, just remove and test the orange drops out of circuit. They should not read as low value resistors.

I got a couple of these and they tell me the ESR and resistance and capacitance of any common capacitor. Get one or two.

https://www.ebay.com/itm/2754341298...Fol5tKVkq9V//orXKQc1zNbF2juD|tkp:BFBMgtiA19Jh

Ty for the reply. When I remove them, do I need a specific meter to test for resistance?

 

[woodbutcher65]

How about the one I linked to? Works great for basic value tests including ESR.
Otherwise you'd want a capacitor meter that tests ESR.

 

[Jeffw5555]

FYI:

- ESR is not something you test for on bypass and coupling capacitors (film capacitors like orange drops and ceramic capacitors)
- Most of those cheapie test meters are not very good for testing ESR anyway
- ESR is only tested on electrolytic filter caps, because an electrolytic cap can test just fine for capacitance but have a very high ESR, meaning it will not be effective as a filter for AC
- More important for bypass/coupling caps (especially for this discussion) is dielectric leakage, which usually is tested out of circuit using a DMM set to resistance/ohms mode
- However, please note that DMM's use a very low voltage to inject into the part being tested. Many leaky caps used on tube amps will read fine on a DMM but will leak in the presence of high voltage
- For that reason, companies that care to help their techs generally will provide test voltages on their schematics to help quickly zero in on where there may be issues with leaky caps. Mesa is not one of those companies. (they don't even supply good schematics to their techs)

I have a home-built tester for ESR testing of filter caps. It is nothing more than a square wave pulse generator with 1 KHz and 10 KHz settings, injects the pulses into the caps thru a resistor and observing the result on a scope. You should expect and integral waveform, you simply test a good cap, observe the shape, test the suspect cap, and it instantly shows if the suspect cap is bad.

As noted above, coupling/bypass caps can leak at high voltages. It is often easy to isolate the suspect cap and connect one side to high voltage (hundreds of volts DC to one side, connect the other side to a high resistance and measure the other side.) This can be done in circuit on many amps by isolating at least one side of the cap to inject a high voltage. (Since this can be dangerous, I don't advise anyone doing this unless you really know what you are doing)

 

[woodbutcher65]

I've found that really bad coupling caps read as shorts to a regular multimeter. No high voltage needed to tickle them into shorting.
Come to think of it, I haven't yet found a coupling cap that was shorted that didn't read bad on the DMM.

Those cheap little multi testers are very handy. They're not a replacement for a good DMM but when the component is removed from the circuit, they're quite handy at telling you what it actually is.

 

[Jeffw5555]

Yes, if they are really bad, an ordinary DMM can easily find them. However, there are many situations where leaky doesn't mean low resistance, especially using lower voltages. Your cheap tester won't help you find those. The methods I talked about will.

 

[Henz]

How about the one I linked to? Works great for basic value tests including ESR.
Otherwise you'd want a capacitor meter that tests ESR.

Dang, ty. That link was hiding in plain site!!!

 

[Henz]

How about the one I linked to? Works great for basic value tests including ESR.
Otherwise you'd want a capacitor meter that tests ESR.

wood butcher , I'm shot gunning parts of my amps circuitry and why are .02uf's 715 's 400v so hard to find!?? Youd think mouser would have everything I need! Jeesh! Any other shops that you can point me to? I need some ceramics as well. Ty for your time! Jim

 

[woodbutcher65]

I use Panasonic ECW series caps. They're great. Nothing against orange drops, but put in Panasonics and you'll never know the difference. If anything, the Panasonic caps are more reliable.

I view Panasonics as my baseline bypass and coupling caps. I keep them in values for every amp I'm likely to encounter.
I also have other types, such as blue Xiconx, some silver micas, and orange drops.

 

[Henz]

I use Panasonic ECW series caps. They're great. Nothing against orange drops, but put in Panasonics and you'll never know the difference. If anything, the Panasonic caps are more reliable.

I view Panasonics as my baseline bypass and coupling caps. I keep them in values for every amp I'm likely to encounter.
I also have other types, such as blue Xiconx, some silver micas, and orange drops.

Ty! Hey, off topic but as I was lifting and measuring components, that 10 or 20 pf cap in parallel with the 3.3 meg , mine read 30 pf. Now the schematic is kinda misleading with Simul Class. It says to put 20pf with later SC's. Mine has 10pf markings on it. Last time I talked with Mike B. , he told me that the 10 pf is part of the bright reduction mod in SC's. Along with the 1000pf in V4 B ( parallel with the 270k plate). Would there be a big difference between 30 pf and 10 pf? Its supposed to be 10. Thanks for your insight and time!

 

[woodbutcher65]

I'm not going to offer opinions on how component changes will alter the tone of the amp. I would always try it and decide for myself.

 

[Jeffw5555]

Ty! Hey, off topic but as I was lifting and measuring components, that 10 or 20 pf cap in parallel with the 3.3 meg , mine read 30 pf. Now the schematic is kinda misleading with Simul Class. It says to put 20pf with later SC's. Mine has 10pf markings on it. Last time I talked with Mike B. , he told me that the 10 pf is part of the bright reduction mod in SC's. Along with the 1000pf in V4 B ( parallel with the 270k plate). Would there be a big difference between 30 pf and 10 pf? Its supposed to be 10. Thanks for your insight and time!

You can’t read low pf caps with a cheap meter.

 

[woodbutcher65]

I've found these little component testers to be remarkably accurate, in fact, mine are apparently accurate to about 2 percent, based on measurements of some ultra high precision components I have, such as a 11,993 ohm resistor with a tolerance of .05%.
They're certainly very good at measuring capacitors in the pf range. In fact I'd consider them to be ideal and within their intended range anywhere below about 10 microfarads.

 

[Jeffw5555]

I've found these little component testers to be remarkably accurate, in fact, mine are apparently accurate to about 2 percent, based on measurements of some ultra high precision components I have, such as a 11,993 ohm resistor with a tolerance of .05%.
They're certainly very good at measuring capacitors in the pf range. In fact I'd consider them to be ideal and within their intended range anywhere below about 10 microfarads.

Nonsense. Almost all cheap meters cannot accurately measure a 10 pf cap.

 

[woodbutcher65]

Believe what you want. I'll believe the results I'm getting based on the evidence right in front of my face..

 

[Henz]

The meter that i have is fairly precise with larger caps. So ill hope and assume its ok for smaller ones. I dunno. Let me ask though if I may. I m pretty sure my meter will test for leaky DC voltages in circuit. This is how I do it. Please let me know if I'm doing it wrong. an example as to how I measure for leaky caps. Let me take the tone stack for example. I attach the ground probe to chassis. Plate gives off the high DC voltage to one side of the .1 and .047 . My meter reads this with the positive probe. Then I take positive probe and test other side of orange drop. They usually read .5 or so mv. My ground probe is still attached to chassis. Am I doing it correctly? With my DRG C+, my Volume 1 pot , while increasing, it acts normal. But when it approaches 7 to 8 or so, buzz starts to present itself. Buzz disappears when pot is at 10. Any ideas as to what this maybe? Tia! Jim.