Fender 400PS Bias issues

[Scott MacKenzie]

Hello all

New member here. I have been repairing tube amps for some time and understand the dangers and safety issues involved with working with high B+ voltages.

My question here is how to bias the 400PS correctly. I have been following the Ed Jahns 30 point setup

The Fender 400-PS Updates & Mods

I can get the -39VDC on the bias pots and all components were verified to be installed correctly. I have a measured resistance across OT at 6.2 ohms and 5.5 per side. 692 & 695VDC per side plate voltage, but voltage drop of 0.519 & 0.304VDC per side. This is giving me 53.6 & 35.2 watts plate dissipation which I think is scary high for 6550's. I installed new electro caps throughout the amp, all to spec.

What is the ideal plate dissipation for this amp, how can I balance the voltage drop of the OT, and does a radically different PD effect the performance? Schematic is in the link above.

Thanks to any knowledgeable minds who can point me in the right direction.

Cheers
Scott

 

[Wasnex]

Hello all

New member here. I have been repairing tube amps for some time and understand the dangers and safety issues involved with working with high B+ voltages.

My question here is how to bias the 400PS correctly. I have been following the Ed Jahns 30 point setup

The Fender 400-PS Updates & Mods

I can get the -39VDC on the bias pots and all components were verified to be installed correctly. I have a measured resistance across OT at 6.2 ohms and 5.5 per side. 692 & 695VDC per side plate voltage, but voltage drop of 0.519 & 0.304VDC per side. This is giving me 53.6 & 35.2 watts plate dissipation which I think is scary high for 6550's. I installed new electro caps throughout the amp, all to spec.

What is the ideal plate dissipation for this amp, how can I balance the voltage drop of the OT, and does a radically different PD effect the performance? Schematic is in the link above.

Thanks to any knowledgeable minds who can point me in the right direction.

Cheers
Scott

Here's my guess. The schematic says Bias set for Ip total 210mA. This averages out the 35mA per tube.

 

[byacey]

I believe the bias should be much higher than that, somewhere between -53 to -60 volts. -39 seems way too low considering the plate voltage involved. The schematic looks to have a glitch where the bias voltage is shown; perhaps -59V?

My mistake, I see they are using cathode bias too, along with fixed bias; kind of a kooky circuit.

 

[Scott MacKenzie]

Thanks for the replies. It is an unusual circuit.

 

[Wasnex]

I don't have any experience with this amp so take everything I write with a grain of salt.

Note in the procedure in the link you provide: When set to operate with no speakers connected, the amp should pull 1.5-1.7A. With one speaker connected, the current should be 2.1-2.5A. With all three speakers connected, the current should be 3.2-3.4A. This is useful info if you have means to measure how much current the amp is pulling.

Although I am not certain, my perception is the secondary of the OT and switching of the output jacks is not drawn accurately. If I understand the circuit, the output jacks switch the tubes on and off in pairs. Each tube has a 10 ohm resistor connected to the cathode, so you can calculate the current each tube is pulling by measuring the voltage drop across the 10 resistors. Basically 350mV indicates 35mA. The part that is not clear to me is whether the 10 ohm resistors are effectively grounded when a speaker plug is inserted in one of the output jacks; I suspect it is.

I would use the effective plate voltage to calculate 70% dissipation and keep the total current for each tube below that figure. Effective plate voltage is the voltage between the plate and cathode.

Rather than doing calculations, I just punch the plate voltage and tube type into this online calculator. Tube Bias Calc Assuming the effective plate voltage is 690V and a 35W tube rating, 70% is 35mA per tube, which matches up with what the schematic says.

Of course that's total current and some of the current will go to the screens. You could measure the voltage drop across the screen resistors and derive the screen current for each tube if you wanted to get really picky. I usually just set the cathode current around 70% and call it a day...unless of course the amp has specific procedures that specify a different bias range, like SVTs which are biased fairly cold. FYI, SVTs also have around 690V on the plates but target current is 72/3=24mA

So basically set the bias and balance so all the tubes are idling around 35mA. Reshuffle the tubes as needed to get the best current match.

After your done, set the hum balance for minimum hum.

The procedure in the page you linked says to set the bias to -39V. That will probably be way too low IMHO. I would turn the bias to a much higher value and then set the bias to produce a safe amount of tube current. FYI, the schematic for vintage SVTS shows bias at -47V.

The article you linked references GE 6550A which are 42W tubes. I would still go with the 35W ratings especially since notes on the schematic suggest the target current is 35mA.

Good luck.

 

[byacey]

It appears plugging into the appropriate speaker jack grounds the 10 0hm cathode resistors, otherwise they are connected to ground through the 4.7K bleeder to prevent the cathodes from floating up to a potentially high voltage when there isn't a plug in the jack.

 

[okcrum]

The 400PS is the most unusual design Fender has issued. Here's a description of dealings with the designer, Ed Jahns, by a service tech. I think you'll understand.

 

[Wasnex]

The 400PS is the most unusual design Fender has issued. Here's a description of dealings with the designer, Ed Jahns, by a service tech. I think you'll understand.

My variac has a built in ammeter, so I could perform the line current procedure. But IMHO measuring the voltage drop across the cathode resistors is the way to go. This will tell you the current pulled by each tube. There will be a bit of error because the resistors are not perfectly matched, but I still think it will be closer than the line current method.

The article's claim that the line current method is the only way to set up the amp is IMHO BS. Funny how they try to explain the procedure as difficult to comprehend or mysterious. It's actually very basic knowledge as long as the amp's current specs are known.

In fact, I use the line current method on some of my amps that do not have bias sense resistors. First I use bias probes to ensure the current is balanced between tube and in the correct range. When I get the amp dialed into the proper range, I check the line current on my variac. Unfortunately the bias probes skew the settings slightly, so after I pull the bias probes out of the amp, I need to reset the bias to reestablish the proper line current reading.

The warning that the setup procedure is specific to the GE 6550A, could be a good indicator to run the tubes a bit cooler than 70%. I don't know if it's true or not, but the amp seems to have a reputation for being hard on tubes, so it might be a good idea to experiment and see how the amp sounds if you cool the tubes down go around 30mA.

 

[okcrum]

I remember the thing about these amps and GE tubes, but the [Invalid or Expired Link Removed] I have calls for selected Tung Sol output tubes. So it seems there were a few different approaches to this amp.

At 700V I can see how it would be hard on tubes rated for 660. KT88s (800V) would have been better.

 

[_One//Man//Mutiny_]

I don't have any experience with biasing the tubes as mine is still rocking the tubes it's had since I bought it. Commenting because it's few and far between there's mention of this amp anywhere and I want to follow along!

 

[okcrum]

At 700V I can see how it would be hard on tubes rated for 660. KT88s (800V) would have been better.

I forgot to add that some modern 6550s (like JJ) are only rated to 600V.

 

[Scott MacKenzie]

Hey All.

Thanks for all the input. So, to summarise, I was a dumbarse....I tried balancing the tubes using the hum pot. Once I connecter 2 DMM's to both sides of the tubes could I balance the tubes correctly. Then, using Wasnex' info, I connected 2 DMM's across the pair of tubes I was measuring. They biassed to 35MV beautifully. I put Tungsol sextet into it. I reckon this is a good place to start for this beast. BTW, shag me, it makes the walls shake!!!!! ;-))

 

[byacey]

Instead of using two meters, To check balance between the pair, measure voltage between the two cathodes. A balanced state will yield zero volts between the cathodes.

 

[Wasnex]

I remember the thing about these amps and GE tubes, but the [Invalid or Expired Link Removed] I have calls for selected Tung Sol output tubes. So it seems there were a few different approaches to this amp.

At 700V I can see how it would be hard on tubes rated for 660. KT88s (800V) would have been better.

The images in one of the links from earlier in the thread shows Tung Sols as well. IMHO I have doubts whether modern Tung Sols would hold up, as many of the newer New Sensor 6550 and KT88 have trouble with high plate voltage.

From the reading I have done, another problem with this amp is the output tubes are really too close together. If the amp was mine, I would be tempted to throw a set of JJ KT88 in it. However the glass envelope of KT88s tends to be a wider in diameter than 6550s, so this would probably exacerbate the thermal problems.

The voltages in the amp appear to be about the same as an SVT, and quite a few people seem to have trouble keeping modern 6550s healthy in SVTs. Some users hear have had good luck with current production Tung Sols.

The current production 6550 I would be most likely to try is the Sovtek 6550WE. The datasheet shows a 42W plate dissipation rating, the plate voltage limit is 800V and the screen voltage limit is 440V. You won't find actual datasheets for most of the other New Sensor 6550 and KT88 tubes. The company basically just republishes datasheets from the old tube companies, but the current production tubes really having nothing in common with the vintage tubes. The 6550WE seems to have a fairly good reputation for taking a pounding, but I don't know if data from tube sellers supports common wisdom or not.

 

[Wasnex]

Hey All.

Thanks for all the input. So, to summarise, I was a dumbarse....I tried balancing the tubes using the hum pot. Once I connecter 2 DMM's to both sides of the tubes could I balance the tubes correctly. Then, using Wasnex' info, I connected 2 DMM's across the pair of tubes I was measuring. They biassed to 35MV beautifully. I put Tungsol sextet into it. I reckon this is a good place to start for this beast. BTW, shag me, it makes the walls shake!!!!! ;-))

Per the schematic: This amp uses a transformer with two secondaries instead of a phase inverter tube. One of the secondaries feeds three of the output tubes and is connected in phase. The other secondary feeds three of the output tubes and is connected out of phase. The bias is fed through the secondaries of the transformer to the grid networks from a tapped potentiometer. My assumption is you balance tube current across the PT by adjusting the tapped potentiometer. No idea where this tapped potentiometer resides in the amp.

 

[gbazinet]

Been a lot of talk about these amps here today!