Biasing a JCM900 SLX model 2500

[Kuga]

Hi

It's a 50W EL34 model. A stock one no mods. Acording schematics you can see bias need to be at -41Vdc in R26/R27. But looks like it's a cold bias at -41Vdc because you get 25mV If you mesures voltage across a 1R resistor on pin1. I need about 35mV in a 1R resistor on pin1 to get a 60% dissipation on power valves.

How would be the best way to bias It?

Thanks

 

[Pete Farrington]

Acording schematics you can see bias need to be at -41Vdc in R26/R27. But looks like it's a cold bias at -41Vdc because you get 25mV If you mesures voltage across a 1R resistor on pin1

What's your mains voltage?

If that's not exactly what the designer was expecting, then all schematic voltages will be off to some degree.

Perhaps with the Marshall selected EL34 (Tesla?) used at that time, a nominal -41V bias provided suitable idle conditions?

Whereas a random matched pair of valves, even though matched to each other, may not idle at the level intended.

I need about 35mV in a 1R resistor on pin1 to get a 60% dissipation on power valves.

It's standard procedure surely? eg tweak VR2 until the desired idle condition is achieved.

https://el34world.com/charts/Schematics/files/Marshall/Marshall_jcm900_slx_50w_2500.pdf

 

[Purgasound]

I've had better luck with current probes for the 900 series. Since the cathodes go through some fuses and travel through the PCB to a ground through some diodes there's a bigger discrepancy when using the 1ohm resistor method. You'll read a couple ohms from cathode to ground without a 1ohm resistor even though there should be continuity. Could be off by 10mV or so.

 

[Pete Farrington]

Since the cathodes go through some fuses and travel through the PCB to a ground through some diodes there's a bigger discrepancy when using the 1ohm resistor method. You'll read a couple ohms from cathode to ground without a 1ohm resistor even though there should be continuity. Could be off by 10mV or so.

Just measure across the resistor legs, ignore the chassis.
The V DC across a 1R resistor in series with pin 8 will still sample cathode current.

Whatever, there should be no current flowing via the hum loop breaking network D6//D7//C1//R36, ie between circuit common and chassis safety earth.

1R resistor on pin1

Pin 8 is the cathode. Pin 1 is the suppressor grid.

 

[Kuga]

What's your mains voltage?

If I use a variac I can adjust bias at 240Vac. The mains voltage on JCM900 it's 240Vac but it's not a big difference at 230Vac.

Perhaps with the Marshall selected EL34 (Tesla?) used at that time, a nominal -41V bias provided suitable idle conditions?

Could be. I use a matched NOS EL34 SVETLANA.

You'll read a couple ohms from cathode to ground without a 1ohm resistor even though there should be continuity. Could be off by 10mV or so.

On this amp I read 23 Ohms.

Just measure across the resistor legs, ignore the chassis.
The V DC across a 1R resistor in series with pin 8 will still sample cathode current.

I did It. If on R26/R27 I have -41Vdc then I have 25mVdc across 1R resistor in series with pin8 / pin1.

Pin 8 is the cathode. Pin 1 is the suppressor grid.

Pin 8 and pin 1 are joined.

 

[Kuga]

After different sound test I decided adjust It like show the schematic with mains at 230Vac. It sounds good to my ears in this way.

 

[Spanngitter]

Did you use the Search? https://www.marshallforum.com/threads/proper-bias-jcm-900-sl-x.106645/

 

[Kuga]

Did you use the Search? https://www.marshallforum.com/threads/proper-bias-jcm-900-sl-x.106645/

No I didn't. I took a look.

 

[Kuga]

I have a pair of EL34 TESLA in good condition from a '93 dual reverb. At -41Vdc on R26/R27 I have 33mV across 1R resistor in series with pin8 / pin1. 449Vdc on plate. About 60% dissipation. This means the -41Vdc on R26/R27 it's for EL34 TESLA.

 

[Pete Farrington]

This means the -41Vdc on R26/R27 it's for EL34 TESLA.

My reading of the schematic is that with 240V mains input, using EL34 whose characteristics are around the centre of the bell curve, -41V at the bias supply output provides the intended operating point.

With your 230V mains (actual measured or just assumed?), assuming your EL34 characteristics are close to the type nominal, your screen grid and anode voltages will be above 4% low, hence -41V will be excessive, should be about 4% less, eg -39.4V.

 

[Spanngitter]

I have a pair of EL34 TESLA in good condition from a '93 dual reverb. At -41Vdc on R26/R27 I have 33mV across 1R resistor in series with pin8 / pin1. 449Vdc on plate. About 60% dissipation. This means the -41Vdc on R26/R27 it's for EL34 TESLA.

Ditch these Voltage recommendations on the Bias, they will fit somehow for the tubes the Vendor selected / installed and surely they will cause the amp to run in an area where it ensures that they last at least during the warranty cycle, only a good probe and some P=U*I calculation will tell you the truth.
Running 60-70% dissipation sounds ok for a 2500 as its power stage should remain linear and not really affect the sound formed in the PreAmp section, but there could be a sweet spot be above or below this setting.
I usually target 70%, adjust that in and then play the amp playing with the Bias until it sounds fine to me. Surely I monitor carefully, using a Bias Probe and ensure I am not exceeding the Specification but maybe your amp comes real to life at 45mA or you like it better at 27mA...let your ears decide.