Magnatone 480A Bias + Plate issue

[FPALB]

I have this 480A that just came in that the owner says is red plating after playing for a half hour or so.

I put in some 1ohm resistors to check the cathode currents and put in a bias pot replacing the 47K resistor in the bias circuit.

The funny thing im seeing is that the plate voltage is changing when I adjust the bias - haven't seen this before.

Any ideas here?

100k grid to bias resistors are spot on
Changed the coupling caps

 

[CJ]

Plate voltage will come down as the rubes are biased to draw more current,

What brand tubes? They might run hot when the amp is in the cabinet,

Or there might be something else going on, like a conductive circuit board,

If that is the case, you can stiffen up the bias supply by lowering resistor values and you could also install cathode resistors line 50 ohms that will help regulate current runaway. And you can also lower screen volrage.

 

[FPALB]

Plate voltage will come down as the rubes are biased to draw more current,

What brand tubes? They might run hot when the amp is in the cabinet,

Or there might be something else going on, like a conductive circuit board,

If that is the case, you can stiffen up the bias supply by lowering resistor values and you could also install cathode resistors line 50 ohms that will help regulate current runaway. And you can also lower screen volrage.

Thanks CJ

Electro Harmonix

480A is Point to Point. I did find an intermittent short between the bias point between the 100Ks and one of the screens. That would definitely shoot the bias through the roof. No issues yet.

Is there anything wrong with using the "remote" outputs for dummy loads? Remote connectors on the front of the 480A break the speaker output connections when used.

 

[trobbins]

Episodes of red-plating may have caused some of the output tubes to significantly out-gas. That may upset the bias from abnormal grid leakage, which may or may not recover with time, and move idle levels around over time. And the 5U4 may be aging, which could make an idle bias adjustment more noticeable for B+ voltage change.

There are some subtle bullet proof changes that could be made, but the owner may not want that.

 

[abbey road d enfer]

The funny thing im seeing is that the plate voltage is changing when I adjust the bias

Seems pretty normal to me. By how much?
Ohm's law says that increasing current also increases voltage drop across the primary's DC resistance.

- haven't seen this before.

You didn't look hard enough.

 

[FPALB]

Episodes of red-plating may have caused some of the output tubes to significantly out-gas. That may upset the bias from abnormal grid leakage, which may or may not recover with time, and move idle levels around over time. And the 5U4 may be aging, which could make an idle bias adjustment more noticeable for B+ voltage change.

There are some subtle bullet proof changes that could be made, but the owner may not want that.

Purchased a new old stock 5U4 just in case. thanks

 

[FPALB]

Seems pretty normal to me. By how much?
Ohm's law says that increasing current also increases voltage drop across the primary's DC resistance.

You didn't look hard enough.

It was changing by 40V ish. But that short was really messing a lot of things up, so im considering that issue solved for the moment. I have it running on some dummies and monitoring the current.

"You didn't look hard enough" - I accept this dig, lol. The short was really moving the bias voltage a lot, and I was just monitoring the cathode current so I was missing some real time information. Ive also never pushed the bias voltage that far off (which is what the short was doing) so ive never seen the B+ vary that heavily. Another one for the lab notebook.

Appreciate it yall.

 

[FPALB]

Now im tackling this horrible hum in the reverb circuit. The 480 separates the reverb to just one of the 2 output amplifiers (which only drive the right speaker). It uses a 6DR7 to drive the tank and I believe the makeup is done via a 2N306 GE transistor. I attached another schematic where ive marked spots in red where grounding or cutting the circuit completely kills the hum+signal. The spots in blue result in no change when grounded.

In this circuit the return reverb cabling is the only cable with its shield attached to ground. The send cabling shield grounds when attached to the tank.

The section after the 5uF going to the 2n306 is actually mounted under the chassis, ive attached a picture. That section returns to the internal chassis area pre the .01uF going to the phase inverter.

Other things ive tried with no improvement:
Removing the tube all together
Shorting the foot switch
recapping this section
disconnecting the reverb tank
swapping out the GE for a modern silicon equivalent
moving reverb send and return cabling
replacing the shielded cabling from the point after the 3.3k summing resistors back to the 5uf capacitor
for sanity I even turned off all my equipment and plugged the amp in some where else.

When the reverb control is at its lowest "Depth" the hum sounds like all 60hz, but when at its max it has a bit of distortion on it.

The 5u4GB is also rattling a lot, so I got a replacement already on its way - dont think this would cause any issues, would it?

Ive attached a video, which if it doesn't work here is a google drive with the video uploaded.

 

[abbey road d enfer]

I hear more 120Hz than 60.
Actually, the noise level doesn't change much with the reverb level, which suggests that noise comes from after it.
Sine you'v etried recapping, we can probably rule out dirty rails.
First ground the connection of the two 3.3k resistsors to the foot of the 4.7k, 1.5k and 100uF (at transistor 2N306).
Does the hum stay?
Is hum equal on both speakers?

 

[FPALB]

I hear more 120Hz than 60.
Actually, the noise level doesn't change much with the reverb level, which suggests that noise comes from after it.
Sine you'v etried recapping, we can probably rule out dirty rails.
First ground the connection of the two 3.3k resistsors to the foot of the 4.7k, 1.5k and 100uF (at transistor 2N306).
Does the hum stay?
Is hum equal on both speakers?

I haven't quite discovered the failure analysis skills of hearing 120hz - how does this help track down failures and how is it becoming 120hz? Of course its related to mains voltage, but can you elaborate?

The schematic attachment shows where I tried cutting to find where the hum is emanating from and one of those cuts that removed the hum is directly post the mix of the reverb and clean signal, so I do believe its some where near there. Definitely Pre the 5uF at the base of the 2N306 (this is one of the places I cut) . Im not an expert on this type of reverb boosting, so any assistance here would be great - my guess is that the circuit uses the 6DR7 stage 1 to boost the input and split that between the input of stage 2 of the 6DR7 for Reverb Tank Driving and a parallel dry signal which then goes through the .022uF and the 330K for control with the reverb depth knob and then summing with the wet signal.

Unsure what you mean by "Foot".

The hum only shows up in the right (if looking from the front of the amp) speaker - the reverb output only goes to that speaker. Each speaker has its own dedicated power amp. I caught this part early by just pulling phase inverter tubes to get a general idea of where the hum was coming from.

 

[trobbins]

Ground the 3k3-3k3-5uF node to the gnd node of the 2N306 circuit ('the foot'), ie. the 4k7-1k5-100uF node.

Is the 6973 PP output stage for that channel operating with ok balance (ie. the same cathode currents), and how did you measure that?

Can you record the hum to a music file (ie. via a direct input box, or a soundcard, or a microphone in a quiet room). If so, then use some free software to check the frequency spectrum of that file - it should indicate both 60 and 120Hz hum peaks - perhaps one of the peaks is larger than the other. It is a faultfinding tool.

Faultfinding hum can be progressed by zoning in on the circuit region it originates in, and excluding from attention circuitry that doesn't cause significant hum. The node grounding technique helps to identify 'before' and 'after' circuitry. It works well if you can 'quantify' the level of hum you are experiencing - that could be your ears, but you need to keep the same conditions, which is why some form of meter measurement is to be preferred.

 

[abbey road d enfer]

I haven't quite discovered the failure analysis skills of hearing 120hz - how does this help track down failures and how is it becoming 120hz? Of course its related to mains voltage, but can you elaborate?

120Hz is the result of rectifying 60Hz. Rather than trying to explain, I refer you to wiki https://en.wikipedia.org/wiki/Rectifier

The schematic attachment shows where I tried cutting to find where the hum is emanating from and one of those cuts that removed the hum is directly post the mix of the reverb and clean signal, so I do believe its some where near there. Definitely Pre the 5uF at the base of the 2N306 (this is one of the places I cut) . Im not an expert on this type of reverb boosting, so any assistance here would be great - my guess is that the circuit uses the 6DR7 stage 1 to boost the input and split that between the input of stage 2 of the 6DR7 for Reverb Tank Driving and a parallel dry signal which then goes through the .022uF and the 330K for control with the reverb depth knob and then summing with the wet signal.

Unsure what you mean by "Foot".

The hum only shows up in the right (if looking from the front of the amp) speaker - the reverb output only goes to that speaker. Each speaker has its own dedicated power amp. I caught this part early by just pulling phase inverter tubes to get a general idea of where the hum was coming from.

That confirms that noise comes from the reverb recovery circuit.
I see three possibilities:

• noise comes from the output of reverb tank; this should be easily diagnosed by inserting a floating TRS jack in the Rev footswitch socket. If noise ceases, it's the tank picking up magnetic interference from the power transformer.
• noise comes from dirty supply voltage B. Not very likely, since that would mean most of the caps are defective.
• noise comes from dirty ground at the "foot". That is something I would look into, suspecting dry joint:faulty ground.

 

[CJ]

You can use your phone to generate 60 or 120 Hz , hold it near the amp and you will instantly know what freq the amp is making,

 

[abbey road d enfer]

Actually, the frequency is 180Hz, third harmonic of mains frequency. Seems there is something a little more complex happening.
Unless my smarthone does silly things...
EDIT. Meantime I checked my phone app acuracy. It's perfect.
Later edit: Rechecked my phone app; the waveform in the video is such taht it fools the frequency counter.
I can confirm teh tone is mainly 120Hz, with a lot of harmonics.

 

[FPALB]

Thank you all for your responses - had an onsite all last week so ill be following up with this tomorrow most likely. All the best.

So basically the 120hz from a diagnostic perspective is to check for pre or post rectification noise - that makes sense.

 

[FPALB]

Ground the 3k3-3k3-5uF node to the gnd node of the 2N306 circuit ('the foot'), ie. the 4k7-1k5-100uF node.

Is the 6973 PP output stage for that channel operating with ok balance (ie. the same cathode currents), and how did you measure that?

Can you record the hum to a music file (ie. via a direct input box, or a soundcard, or a microphone in a quiet room). If so, then use some free software to check the frequency spectrum of that file - it should indicate both 60 and 120Hz hum peaks - perhaps one of the peaks is larger than the other. It is a faultfinding tool.

Faultfinding hum can be progressed by zoning in on the circuit region it originates in, and excluding from attention circuitry that doesn't cause significant hum. The node grounding technique helps to identify 'before' and 'after' circuitry. It works well if you can 'quantify' the level of hum you are experiencing - that could be your ears, but you need to keep the same conditions, which is why some form of meter measurement is to be preferred.

Threw a microphone in the back of the cab and used "Signal Scope" software (which I rather like) to view the FFT. This is viewing with the Y axis in Log with no smoothing - these options gave the clearest image.

Image "01" is the noise and its harmonics - some 60 hz, but definitely a larger 120hz

Grounding the 3k3-3k3-5uF node to the gnd node of 2N306 Foot increased the 120hz dramatically, Image "02", (view options changed to actually fit into the screen).

Something to note is that the 4.7k ground, in the 2N306 foot, is soldered into a terminal strip inside of the chassis and the 1.5k and 100u are soldered to a terminal strip thats grounded to the sheet of steel thats mounted under the chassis (picture attached). That sheet of steel is grounded through its mounting screw. I joined these two with a jumper to see if that noise went away and nothing changed. The terminal strip on the sheet of steel contains most of the 2N306 circuitry (the Pot, the 220k, the 1.5k, the 2N306, and the 100u).

6973 cathode currents are both 28mA (~70% of max plate dissapation for a 6973 - 12W max with a plate voltage of ) - I got this by added 1ohm cathode resistors to all the 6973. I added a bias control that controls all bias voltages for all the power tubes.

 

[FPALB]

When probing the speaker itself I get the attached image. Be advised im on channel 2 of my scope. The triggered wave form is at 60hz

 

[FPALB]

AHA!

found a darkened resistor in the power section that I went to change when I discovered a filter cap that had its lead fractured (just barely) by the magnet of the amps tweeter (really narrow margin of chassis insertion). I think this might have happened by my own hand without realized it. I am pretty sure we found the source everyone. Bad part is I dont have it in stock, lol. Ordering it now.

Ill follow up when it gets installed.
All the love to you geniuses.

 

[FPALB]

That confirms that noise comes from the reverb recovery circuit.

The recovery circuit being the 2N306 circuit right? Im confused how this could be since I cut right after the 3.3k resistors and before the 5uF and the hum stopped - wouldn't this mean the source of the issue is before the 3.3ks?


Somehow....replacing this damaged cap and resistor made no difference. The resistor was from some intermittent shorting from a glob of solder that ive fixed.

I realized that Rail B and Rail A (the rail thats at the anode of the reverb tubes) are almost the same voltage, so to rule out the rail I powered the anodes of the reverb section with Rail B. This stressed the voltage divider resistors in the power supply a little bit but they managed through it. This made zero difference which really bummed me out as I was really beginning to think it was a power supply issue.

I also removed the Rail (A) supply all together from the reverb section and it made zero difference....so not the supply I think?

Im super stumped on this issue.

 

[abbey road d enfer]

The recovery circuit being the 2N306 circuit right? Im confused how this could be since I cut right after the 3.3k resistors and before the 5uF and the hum stopped - wouldn't this mean the source of the issue is before the 3.3ks?

The reverb recovery circuit starts with the reverb tank output. Have you tried to short or disconnect it?