Marshall JMP 2203 - 6550 Bias

[Cap217]

I have an 80 jmp that is setup for 6550s. When I got it, it had el34 with the bias pot maxed abs getting 12ma. So I decided to try 6550 because I don’t have any amps with them. I assumed I would put these in and bias up really well. But that’s not happening. Now I can’t bias cold enough.

From reading I hear bias as a 25w tube or bias as a 35w tube at 60%.

My plate voltage is 500. I can only get these down to 45ma which is 90% of a 25w but under 69% of a 6550. 60% would be 50ma. This is all at 116v wall. 117v gets me the same bias but 505 plate voltage. 120v wall gets me 525 plate voltage and 49ma bias.

So what do I do here? Am I ok or should I make some bias resistor adjustments?

 

[Mikeyboyeee]

How does it sound?

 

[Rick Lee]

525VDC plate sounds high to me. I don't know that any of my Marshalls are over 479VDC and my wall voltage is usually 123VAC. The Eurotubes calculator shows 6550s to be 35w. So you should be good, if not a little cold. Though, given the current tube situation, you may want to keep them cold a while longer.

 

[glpg80]

Plate voltage should be monitored as bias is adjusted. You need to let the tubes warm up for an hour before biasing.

As plate current increases, plate voltage will drop. You have to calculate idle dissipation from plate current and plate voltage both.

My tung sol 6550s are setup at 46mA at 463V which is well below their rates capability and assures long life.

 

[Cap217]

How does it sound?

Sounds good so I think I’m going to keep it as is.

Plate voltage seems high but Have others that are around the same. Then others at 470. Then a 50w at 350.

 

[harddriver]

I forgot that 6550's were originally rated for 42 watts dissipation sorry my brain was thinking 6CA7(35 watt) so I amended my post.

Any new production 6550 I usually bias it as if it is a 35watt dissipation even though they may claim they are 42 watt. That will help to not stress the tubes too much. You must increase the negative bias range to run the 6550's in their bias range. I also like to bias my 6550's cold at 50%-60% tops. 6550's lose their definition and get smeared sounding at 70% dissipation. My SED6550's have been running in my 2203 for years now and they keep on chugging with no issues. Original GE6550's are also bulletproof and are true 42 watt dissipation. You may have to change one or more of the bias feed resistor to increase negative bias and get it into the bias range listed below. I love my 6550's as I have tried every EL34 and variant out there but YMMV. Your B+ voltage sounds kind of high for a 1980 Marshall with the stock power transformer. My all original 1979 2203 had 454DCV plate voltage, it's not impossible that yours is 500DCV but i have not seen that many from 78-80 with plate voltages that high. I think starting with the JCM800 models some go into the 500DCV plate voltage ranges. My current 2203 has a NOS stock Marshall dagnall power transformer from the 2000's and it supplies about 460DCV to the plates.

If your amp still has the 220K bias splitter resistors to can run then as long you bias the amp cold but if you are concerned about the 6550 going into runaway bias Marshall used to reduce the bias splitter resistors to 150K you can go as low as 82K but you will have to increase the .022 coupling caps to .047uf or .1uf to get the same bass response but it's not mandatory if you just go down to 150K. The concerns Marshall had with 6550's going into runaway bias never really materialized but they used the 150K resistors just to be on the safe side. My Splawn modded Marshall 2203 is running 220K and I have never had any issues but I have used 82 K on other Marshall builds.

Here's are safe bias ranges to work within safely: From Dan Torres book Inside Tube amps. These are the bias voltage measured in DCV negative voltage method measured on Pin 5 of you power tube socket.
Marshall 100 watt amps EL34's -36 to -45 Recommended -42
Marshall 50 watt amps EL34's -34 to -40 Recommended -36

Marshall 100 watt amps 6550 -44 to -55 Recommended -52
Marshall 50 watt amps 6550 -38 to -48 Recommended -44

 

[LP Freak]

Plate voltage should be monitored as bias is adjusted. You need to let the tubes warm up for an hour before biasing.

As plate current increases, plate voltage will drop. You have to calculate idle dissipation from plate current and plate voltage both.

My tung sol 6550s are setup at 46mA at 463V which is well below their rates capability and assures long life.

An hour? ?

 

[DanTravis62]

Man, 500 at the plate seems kind of high for that era of JMP, I've never seen anything above 470 or 480 at absolute tops.

But if it sounds good, it is good.

 

[glpg80]

An hour? ?

Do you want your bias to drift with shit tubes and run away or not?

My amp has the 150k bias splitters supporting 6550s, but even then, there’s drift until about an hour due to tube quality these days. Actual data sheets show an hour of warmup time and I’ve noticed you need that time else you risk drift.

 

[Cap217]

Any new production 6550 I usually bias it as if it is a 25watt dissipation even though they claim they are 35 watt. That will help to not stress the tubes too much. You must increase the negative bias range to run the 6550's in their bias range. I also like to bias my 6550's cold at 50%-60% tops. 6550's lose their definition and get smeared sounding at 70% dissipation. My SED6550's have been running in my 2203 for years now and they keep on chugging with no issues. Original GE6550's are also bulletproof and are true 35 watt dissipation. You may have to change one or more of the bias feed resistor to increase negative bias and get it into the bias range listed below. I love my 6550's as I have tried every EL34 and variant out there but YMMV. Your B+ voltage sounds kind of high for a 1980 Marshall with the stock power transformer. My all original 1979 2203 had 454DCV plate voltage, it's not impossible that yours is 500DCV but i have not seen that many from 78-80 with plate voltages that high. I think starting with the JCM800 models some go into the 500DCV plate voltage ranges. My current 2203 has a NOS stock Marshall dagnall power transformer from the 2000's and it supplies about 460DCV to the plates.

If your amp still has the 220K bias splitter resistors to can run then as long you bias the amp cold but if you are concerned about the 6550 going into runaway bias Marshall used to reduce the bias splitter resistors to 150K you can go as low as 82K but you will have to increase the .022 coupling caps to .047uf or .1uf to get the same bass response but it's not mandatory if you just go down to 150K. The concerns Marshall had with 6550's going into runaway bias never really materialized but they used the 150K resistors just to be on the safe side. My Splawn modded Marshall 2203 is running 220K and I have never had any issues but I have used 82 K on other Marshall builds.

Here's are safe bias ranges to work within safely: From Dan Torres book Inside Tube amps. These are the bias voltage measured in DCV negative voltage method measured on Pin 5 of you power tube socket.
Marshall 100 watt amps EL34's -36 to -45 Recommended -42
Marshall 50 watt amps EL34's -34 to -40 Recommended -36

Marshall 100 watt amps 6550 -44 to -55 Recommended -52
Marshall 50 watt amps 6550 -38 to -48 Recommended -44

Good info. But my amps plate voltage is so different between amps.

What would you bias a 74 super lead 50w. El 34. 350 plate voltages at 117v wall?

 

[harddriver]

Good info. But my amps plate voltage is so different between amps.

What would you bias a 74 super lead 50w. El 34. 350 plate voltages at 117v wall?

You can still set your bias according to your ma current draw of each tube. With the power tubes out of the amp you would want to sweep your bias pot from minimum and maximum setting and record the bias range for 6550. When your bias supply resistord are correct values for 6550 you should be in the specified bias range as stated in above ranges, then you know you are safe to install you new tubes.

With your multimeter set for DCV on pin 5 of you power tube socket your total sweep of the bias pot should give you at least this much adjustment range, if not you will need to either swap out at least one of you bias supply resistors to obtain the range specified. Once you have this acceptable adjustment range you can then fine tune the bias to your liking according to ma current draw if you have 1 ohm resistors installed on pins 9-1 to ground which it sounds like have or if you are using something like a bias right digital bias tool that either reads in DCmv or DCma.

It's always a good idea to turn your bias pot all the way down when installing a set of tubes where you don't know how much current they will pull when you apply full plate voltage on first fire up so you don't burn them up.

More - bias setting results in more current ma draw resulting in a hotter bias setting and hotter running tubes.

For your 74 50 watter with 6550's this is bias range you are looking for prior to installing the tubes from minmum to maximum sweep of your 25K bias pot.

Marshall 50 watt amps 6550 -38 to -48 Recommended -44

Most tube bias drift upon initial start up will settle in within 15-20 minutes. You then play the amp at volume for 30-60 minutes then re-check the bias and see if it has drifted at all usually the tube tend to drift upward pulling more current. You may need to turn the bias down to address the drift then you should be good to go.

 

[Cap217]

You can still set your bias according to your ma current draw of each tube. With the power tubes out of the amp you would want to sweep your bias pot from minimum and maximum setting and record the bias range for 6550. When your bias supply resistord are correct values for 6550 you should be in the specified bias range as stated in above ranges, then you know you are safe to install you new tubes.

With your multimeter set for DCV on pin 5 of you power tube socket your total sweep of the bias pot should give you at least this much adjustment range, if not you will need to either swap out at least one of you bias supply resistors to obtain the range specified. Once you have this acceptable adjustment range you can then fine tune the bias to your liking according to ma current draw if you have 1 ohm resistors installed on pins 9-1 to ground which it sounds like have or if you are using something like a bias right digital bias tool that either reads in DCmv or DCma.

It's always a good idea to turn your bias pot all the way down when installing a set of tubes where you don't know how much current they will pull when you apply full plate voltage on first fire up so you don't burn them up.

More - bias setting results in more current ma draw resulting in a hotter bias setting and hotter running tubes.

For your 74 50 watter with 6550's this is bias range you are looking for prior to installing the tubes from minmum to maximum sweep of your 25K bias pot.

Marshall 50 watt amps 6550 -38 to -48 Recommended -44

Most tube bias drift upon initial start up will settle in within 15-20 minutes. You then play the amp at volume for 30-60 minutes then re-check the bias and see if it has drifted at all usually the tube tend to drift upward pulling more current. You may need to turn the bias down to address the drift then you should be good to go.

My 74 is el34. Your info says 34-40. That would be fine if the amp had a 450-470 plate voltage but when the plate voltage is 350 don’t you change? I biased to 45ma which is about 65%. But according to the data you provided I am too high.

 

[blackba]

Just serviced a 79 jmp 2203, which had original 6550s that still tested good. At 123vac line voltage, plate voltage was 463vdc. Bias current was 51, 52, 40, and 40mA. So not the most matched set, but they sounded great.

New production 6550s I would bias in el34 range as 25w tubes. I also get tubes from suppliers who do a 24 hour burn in, so I don’t wait for the amp to warm up an hour before setting bias. I did check a few after that long and have never had drift, but as always ymmv.

Also wondering what the plate voltage was with the el34s biased at 12mA, must have been high based on what is being reported with the 6550s.

 

[harddriver]

My 74 is el34. Your info says 34-40. That would be fine if the amp had a 450-470 plate voltage but when the plate voltage is 350 don’t you change? I biased to 45ma which is about 65%. But according to the data you provided I am too high.

Ok I understand your question now. Your Plate Voltage has no bearing whatsover on you bias range, as I said your bias range should read -34 to -40 for EL34's in a 50 watter. You use your pate voltage to determine how much current draw in an amp versus the tube wattage dissipation versus plate voltage.

So with your power tubes completely removed from the amp your bias range should be as follows for a EL34 Marshall. You will need a Digital Multimeter set to read DCV to read your bias voltage at pin 5 on your power tube sockets, you are not reading current draw on pin 5 it's negative bias voltage. Bias voltage and tube current draw in milliamps are two completely different things so try not to get them confused.

Marshall 50 watt amps EL34's -34 to -40 Recommended -36 THIS is DCV bias voltage NOT ma current draw of the tubes

Once you know you you have the correct range you can then calculate your % of dissipation current draw to fine tune to you liking more or less ma current draw. With more - bias your tube will draw more current ie: more ma.

If you do not have a digital multimeter then you will have to rely on only milliamp current draw which I don't recommend. Once you know that you have the acceptable bias range available you can then fine tune using the current (ma) method.

 

[SpiderWars]

I recently went thru this with an '80 JMP. It idles at about 460vdc. There are some that say the current production 6550 is really constructed as a 42W tube and the Tung Sol RI data sheet says 42W plate dissipation. So the tube itself will probably be fine with whatever you choose. But you probably don't want your amp dissipating 70% x 42W/tube x 4 tubes=117.6W at idle either.

Based on your numbers it looks like it's set up for 6550s but you have a high B+ so it's running hot. It's a little easier to get EL34s to bias up because you can just add a parallel resistor to an existing one in your bias supply but then you'll have the 150k splitters which do sound different with EL34s...but you might like it.

My '80 came with Tung Sol RI 6550 running at 61mA per tube at 455vdc. When I lowered it to 65% x 25W/460vdc it did not sound as good. I increased it to somewhere between 65% of 30W and 35W.

All of that to say...if it's just running a little hot relative to EL34s then it's probably fine. Check your PT temp, mine stays cool if it's just idling.

 

[Cap217]

Just serviced a 79 jmp 2203, which had original 6550s that still tested good. At 123vac line voltage, plate voltage was 463vdc. Bias current was 51, 52, 40, and 40mA. So not the most matched set, but they sounded great.

New production 6550s I would bias in el34 range as 25w tubes. I also get tubes from suppliers who do a 24 hour burn in, so I don’t wait for the amp to warm up an hour before setting bias. I did check a few after that long and have never had drift, but as always ymmv.

Also wondering what the plate voltage was with the el34s biased at 12mA, must have been high based on what is being reported with the 6550s.

520 plate voltage with cold el34.

 

[Cap217]

Ok I understand your question now. Your Plate Voltage has no bearing whatsover on you bias range, as I said your bias range should read -34 to -40 for EL34's in a 50 watter. You use your pate voltage to determine how much current draw in an amp versus the tube wattage dissipation versus plate voltage.

So with your power tubes completely removed from the amp your bias range should be as follows for a EL34 Marshall. You will need a Digital Multimeter set to read DCV to read your bias voltage at pin 5 on your power tube sockets, you are not reading current draw on pin 5 it's negative bias voltage. Bias voltage and tube current draw in milliamps are two completely different things so try not to get them confused.

Marshall 50 watt amps EL34's -34 to -40 Recommended -36 THIS is DCV bias voltage NOT ma current draw of the tubes

Once you know you you have the correct range you can then calculate your % of dissipation current draw to fine tune to you liking more or less ma current draw. With more - bias your tube will draw more current ie: more ma.

If you do not have a digital multimeter then you will have to rely on only milliamp current draw which I don't recommend. Once you know that you have the acceptable bias range available you can then fine tune using the current (ma) method.

Ok. I completely misunderstood what you were saying. You are just talking about stock ranges of the bias pot. I understand now.

My bias range for 6550 is 45 - I think 70 on the high side.

 

[blackba]

520 plate voltage with cold el34.

Just to be clear, you are measuring from pin 3 on the power tube to ground? What are you using to measure the plat voltage? Any idea what your line voltage is, as that will affect plate voltage (and all the voltages in the amp)?

 

[Cap217]

Just to be clear, you are measuring from pin 3 on the power tube to ground? What are you using to measure the plat voltage? Any idea what your line voltage is, as that will affect plate voltage (and all the voltages in the amp)?

I am using a weber bias probe for all 4 tubes and a eurotube that does 1 at a time.

My post said 116 wall voltage. It did jump to 117v but that shouldnt matter. I use a brown box. Wall voltage 123v and I reduced it. I usually run my amps at 117v when I play.

 

[blackba]

I am using a weber bias probe for all 4 tubes and a eurotube that does 1 at a time.

My post said 116 wall voltage. It did jump to 117v but that shouldnt matter. I use a brown box. Wall voltage 123v and I reduced it. I usually run my amps at 117v when I play.

Ouch, your plate voltage should not be that high at 116V to 117VAC coming in. Sorry missed where you mentioned this at the original post. something is a bit of a amiss here, IMO