How does this equate to overdrive/distortion, seems to be a fairly similar power rating to the EL34 and 6L6 and the marshalls can deliver 100 clean watts.
Yes I was going by the sound when i adjusted the negative feedback with the pot, to be honest, the marshall amps have negligible effect with its presence controls, i got this working so that with presence set to zero that its got no effect on sound but as I adjust it, i can hear changes, when I put it on the scope, it seems to confirm this too, so my ears must be right.
about 70W, which I feel is too low
They didnt seem to indicate that 50mA was too much, but I dont see much difference in power by making the adjustments until I get closer to -50V bias, so i expect that biasing by the crossover distortion on the scope must be fairly accurate then? I will need to measure the power output and see what difference this makes. Going by current, and not scope readings, i have reduced more bias than where the crossover distortion vanished.
They suggest I go back with the marshall ECC83 phase inverter, this sounds like something I will explore, Im not too sure if any major modifications need to be made with the use of 807 tubes.
Doesn't look far off with a B+ of only 400V (see my post#18).
100W Marshalls have a B+ between 450 and 500V.
Last edited:
Ok, I might not worry too much about that then.
So If that's the case, nothing I do with the phase inverter will make any major impact?
So If that's the case, nothing I do with the phase inverter will make any major impact?
I am looking at rebiasing this amp tomorro, what target should I aim for per tube, is 40mA more acceptable? Will confirm what power draw i am getting when i scope it for crossover distortion
I've got the tubes biased around 40mA each and there is little difference in power output.
Now i have my signal generator set to 150mV @1KHz and the speaker output reads 28V which is 98W on myb8 ohm load.
If i feed in a higher signal it doesn't clip until around 170mV, see photos attached.
The guys on the marshall forum are telling me that the marshall design should only require 10mV to reach full clean power output.
I'm stoked that it's giving me 100 clean watts, but they are telling me that the phase inverter is to blame in this regard, i can only get about half a watt if i feed this with a 10mV signal.
I also have both volume knobs on full to reach full power on the low gain.
Now i have my signal generator set to 150mV @1KHz and the speaker output reads 28V which is 98W on myb8 ohm load.
If i feed in a higher signal it doesn't clip until around 170mV, see photos attached.
The guys on the marshall forum are telling me that the marshall design should only require 10mV to reach full clean power output.
I'm stoked that it's giving me 100 clean watts, but they are telling me that the phase inverter is to blame in this regard, i can only get about half a watt if i feed this with a 10mV signal.
I also have both volume knobs on full to reach full power on the low gain.
Attachments
I guess the 10mV sensitivity from the marshallforum means 10mVrms corresponding to 28mVpp.
I expect the typical Marshall power amp using a 12AX7 PI and EL34 power tubes to have 4 or 5 times more gain than your amp.
I see 3 options:
1) Leave the amp as it is as you seem to like it.
2) Increase the series NB resistor for more gain but less presence. (Leo Fender knew that a 12AT7/6L6 power amp doesn't allow for effective presence boost.)
3) Convert the PI to the 12AX7 Marshall circuit for more open loop gain and also increase the feedback resistor for more closed loop gain and moderate presence boost.
I expect the typical Marshall power amp using a 12AX7 PI and EL34 power tubes to have 4 or 5 times more gain than your amp.
I see 3 options:
1) Leave the amp as it is as you seem to like it.
2) Increase the series NB resistor for more gain but less presence. (Leo Fender knew that a 12AT7/6L6 power amp doesn't allow for effective presence boost.)
3) Convert the PI to the 12AX7 Marshall circuit for more open loop gain and also increase the feedback resistor for more closed loop gain and moderate presence boost.
The Marshall's never rally had much effective presence boost to speak of either.
How ive got this set up is giving me a somewhat noticeable presence boost that's also clearly visible on the scope, certainly much more than a marshall, but at the same time it appears to cut off completely when dialed to zero.
Perhaps I should just back off the NFB a bit more and see how that goes.
I dont want to mess around with the phase inverter just yet, but if I can get some appreciable gain from reducing the NFB at the expense of presence boost, I would consider it.
I can also get close to 100W of fairly clean power on the high sensitivity if I adjust the volume controls appropriately too.
Something else I think might need attention is the resistor from the base of the LTP to ground.
I originally had only 100 ohms here, which turns out is what the fender amps used, however I was getting major crossover distortion and low power, only about 30W max.
I raised it to 4.7K, but would increasing this resistor give me more gain?
It looks like this goes hand in hand with the NFB and acts as a voltage divider?
Right now it's 4.7K on each side.
How ive got this set up is giving me a somewhat noticeable presence boost that's also clearly visible on the scope, certainly much more than a marshall, but at the same time it appears to cut off completely when dialed to zero.
Perhaps I should just back off the NFB a bit more and see how that goes.
I dont want to mess around with the phase inverter just yet, but if I can get some appreciable gain from reducing the NFB at the expense of presence boost, I would consider it.
I can also get close to 100W of fairly clean power on the high sensitivity if I adjust the volume controls appropriately too.
Something else I think might need attention is the resistor from the base of the LTP to ground.
I originally had only 100 ohms here, which turns out is what the fender amps used, however I was getting major crossover distortion and low power, only about 30W max.
I raised it to 4.7K, but would increasing this resistor give me more gain?
It looks like this goes hand in hand with the NFB and acts as a voltage divider?
Right now it's 4.7K on each side.
One of the best sounding Marshalls ever (IMO) was the early 100W (around '66) using a 12AX7/ECC83 PI and 4xKT66 power tubes.
Used by Hendrix on the "Are you experienced" album and by Clapton in the early Cream days.
I think this is the schematic: https://forum.metropoulos.net/download/file.php?id=3106
(The choke wiring is a known schematic error.)
The KT66 is similar to the 6L6/807.
Used by Hendrix on the "Are you experienced" album and by Clapton in the early Cream days.
I think this is the schematic: https://forum.metropoulos.net/download/file.php?id=3106
(The choke wiring is a known schematic error.)
The KT66 is similar to the 6L6/807.
That looks almost identical to the 2203 phase inverter.
Looks like marshall must have stuck with that design in all their 100W units.
As you say, that's likely to work well with my 807 tubes, so I will probably revisit this and change it out when I'm in the mood.
Couple of questions, it's a bit hard to discern the values on the schematic.
Would the cathode resistor be 4.7k on the phase inverter?
I also see that only one tube on each pair shows a grid stopper resistor.
Does this look correct?
I also expect i would want to increase the value of my grid stopper to 15K as shown here along with the 72K resistors to 220K?
Looks like marshall must have stuck with that design in all their 100W units.
As you say, that's likely to work well with my 807 tubes, so I will probably revisit this and change it out when I'm in the mood.
Couple of questions, it's a bit hard to discern the values on the schematic.
Would the cathode resistor be 4.7k on the phase inverter?
I also see that only one tube on each pair shows a grid stopper resistor.
Does this look correct?
I also expect i would want to increase the value of my grid stopper to 15K as shown here along with the 72K resistors to 220K?
Last edited:
1)The PI cathode resistor should be 470R for correct ECC83 bias.
2) I'd use 4 grid stoppers. Anything from 1.5k to 15k should work fine. Except if the amp tends to oscillate.
In that case best use 4x15k.
3) Regarding the power tube grid leaks:
220k gives more PI gain, 72k is safer for the power tubes.
150k (as used with 6550s) might be a decent compromise.
2) I'd use 4 grid stoppers. Anything from 1.5k to 15k should work fine. Except if the amp tends to oscillate.
In that case best use 4x15k.
3) Regarding the power tube grid leaks:
220k gives more PI gain, 72k is safer for the power tubes.
150k (as used with 6550s) might be a decent compromise.
Ok thats one less resistor to worry about, im already using 470 ohms with the existing phase inverter.
My grid stoppers are all 3.9k and ive never noticed any oscillation.
I will leave the grid leaks as they are and see how it performs.
My grid stoppers are all 3.9k and ive never noticed any oscillation.
I will leave the grid leaks as they are and see how it performs.
Class AB amps don't change their maximum power depending on idle current.
OK, any explanation why the output power was changing while I was adjusting the bias level? There was not a huge difference from dropping 10mA, but it was measurable.
I was getting much lower power if i biased the tubes all the way.
I was getting much lower power if i biased the tubes all the way.
Dropping the current in the power amp means you need more voltage swing to drive them to full power.
If you don’t adjust input signal, you get less power. If you target a specific distortion, a guitar preamp will contribute more to the overall distortion thrn power at a specific distortion will be lower.
If you don’t adjust input signal, you get less power. If you target a specific distortion, a guitar preamp will contribute more to the overall distortion thrn power at a specific distortion will be lower.
Ok, I understand. Once I upgrade this phase inverter i will have much more signal strength to drive these tubes too.