Converting ultralinear to pentode mistake?

[Pochie45566]

Was converting a amplifier to pentode operation from ultralinear. I removed the ultralinear taps, hooked up a resistor value of 1k to each g2 of the el34. Hooked those resistors to my first point of voltage in the amplifier after the power transformer. It makes sound and sounds okay. Except the tube is only drawing 15 ma of current. What should I do? Would it help to lower the g2 resistor? Since plate current and g2 current are related I believe.

I meaure 490 volts from g2 to ground however i measure 0 volts across the resistors i installed for each g2 resistor

[DummyLoad]

assuming that your amp is fixed bias. with fixed bias in UL mode the bias is typically deeper (more negative = less current flow through the valve). it is possible that you'll need to modify the bias supply or simply just make the bias a hotter by making the bias more positive (or less negative, however you wish to transpose). if the amp is cathode bias UL, then you'll likely need to reduce the value of the cathode resistor(s).


note that reflected load impedance is not the same in UL and pentode mode - UL mode typically requires a higher load impedance than pentode mode at the same B+ levels. it'll work, but might not work as well as you're thinking it will.


it would help if we know what the amp is and/or provide a schematic in order to supply more accurate corrective advise. 


--pete

[Pochie45566]

Heres a schematic link. wont let me upload it. It is fixed bias. I have a new non linear output transformer also. much less b+ resistance
https://ibb.co/bJvzxk

[Pochie45566]

assuming that your amp is fixed bias. with fixed bias in UL mode the bias is typically deeper (more negative = less current flow through the valve). it is possible that you'll need to modify the bias supply or simply just make the bias a hotter by making the bias more positive (or less negative, however you wish to transpose). if the amp is cathode bias UL, then you'll likely need to reduce the value of the cathode resistor(s).


note that reflected load impedance is not the same in UL and pentode mode - UL mode typically requires a higher load impedance than pentode mode at the same B+ levels. it'll work, but might not work as well as you're thinking it will.


it would help if we know what the amp is and/or provide a schematic in order to supply more accurate corrective advise. 


--pete

any thoughts?

[tubeswell]

The best way to test the output transformer for determining the plate-to-plate load is a VAC test of the turns ratio of the Pri:Sec windings with a low VAC source hooked up to the secondary and measuring the resulting VAC across the primary. For a multi-tap OT, test the VAC ratio on each set of taps.

[Pochie45566]

The best way to test the output transformer for determining the plate-to-plate load is a VAC test of the turns ratio of the Pri:Sec windings with a low VAC source hooked up to the secondary and measuring the resulting VAC across the primary. For a multi-tap OT, test the VAC ratio on each set of taps.

back to my original question though. How can i get a hotter bias? with fixing whatever needs to be fixed

[tubeswell]

If it were my amp, I'd work out what the reflected load options are with that OT first.

[PRR]

> tube is only drawing 15 ma of current

What was it drawing before?

> I removed the ultralinear taps, hooked up a resistor value of 1k to each g2 of the el34. Hooked those resistors to my first point of voltage in the amplifier after the power transformer

The logical place would be point A, the first node, because the DC voltage will be essentially the same as before.

Point B would give much lower Vg2 due to the 22K passing all the small-bottle current and now Ig2 also.

> How can i get a hotter bias?

The plan shows a trimmer. Does yours have this? What is the power tubes' current at either extreme? (If it goes past 50mA, don't linger there.)

[Pochie45566]

> tube is only drawing 15 ma of current

What was it drawing before?

> I removed the ultralinear taps, hooked up a resistor value of 1k to each g2 of the el34. Hooked those resistors to my first point of voltage in the amplifier after the power transformer

The logical place would be point A, the first node, because the DC voltage will be essentially the same as before.

Point B would give much lower Vg2 due to the 22K passing all the small-bottle current and now Ig2 also.

> How can i get a hotter bias?

The plan shows a trimmer. Does yours have this? What is the power tubes' current at either extreme? (If it goes past 50mA, don't linger there.)

it was drawing 37ma. I have about 400v with g2 connected to B. On el34 spec sheet, when plate current is about 500, screen should be about 400. it is at its extreme right now on the trimmer, pulling 15ma.

[Pochie45566]

> tube is only drawing 15 ma of current

What was it drawing before?

> I removed the ultralinear taps, hooked up a resistor value of 1k to each g2 of the el34. Hooked those resistors to my first point of voltage in the amplifier after the power transformer

The logical place would be point A, the first node, because the DC voltage will be essentially the same as before.

Point B would give much lower Vg2 due to the 22K passing all the small-bottle current and now Ig2 also.

> How can i get a hotter bias?

The plan shows a trimmer. Does yours have this? What is the power tubes' current at either extreme? (If it goes past 50mA, don't linger there.)

you were correct. it jumped up to 33ma. But that is still with the trimmer on maximum. I can just adjust the bias circuit now to get lower negative voltage right? The negative voltaghe minimum is 33ma. would I just up the value of that 1k resistor?

[PRR]

> up the value of that 1k resistor?

Sounds like a plan to try.

OTOH, consider reducing that 22K from A to B. You have roughly doubled the current in it. It would be reasonable to make it half the value. The fast-test way is to tack-solder another 22K, same wattage, across the one there.

[Pochie45566]

> up the value of that 1k resistor?

Sounds like a plan to try.

OTOH, consider reducing that 22K from A to B. You have roughly doubled the current in it. It would be reasonable to make it half the value. The fast-test way is to tack-solder another 22K, same wattage, across the one there.

the one for f.b II? or the 20k between A and B the only 22k is see is for F.b. II. how do you know it was doubled?  how should i go about measuring it also?

[DummyLoad]

attached is the schematic of a laney 60 that is a similar amp to yours.

on yours use a 1K 3W or 5W to replace 20K between NODE A and NODE B. see if bias pot will trim down to -40V at pin 5 6CA7.

check reflected - ascertain what the reflected impedance is - tubeswell posted how to do that. if you have a 7-8K or higher reflected impedance to the 6CA7, you'll want to change the speaker load to 1/2 indicated on the speaker jacks - e.g. 4 ohm speakers connected to the 8 ohm jack. 

--pete 

[DummyLoad]

hack up of your schematic with suggested mods attached.


--pete

[Pochie45566]

hack up of your schematic with suggested mods attached.


--pete

you are so helpful! three questions.

Why a 1k there? would a 10k not be better

the bias pot minimum negative voltage is -39v i measured. What should I do to get it to be a bit higher. maybe -20? with it at -39 i am getting 33ma on the tubes. i want 37ma for a 70% bias.

I am rather new to this and confirm if this is correct. tubewell suggests I hook a low voltage ac current into the secondary taps, 4, 8, 16 ohm, of the OT transformer and measure that from the center tap? and find the ratio between the voltage in and voltage out? i know the center tap resistance is about 3.6 which is close the original at 4.3k

http://www.classictone.net/40-18025.html - my ot

[DummyLoad]

hack up of your schematic with suggested mods attached.


--pete

you are so helpful! three questions.

Why a 1k there? would a 10k not be better

the bias pot minimum negative voltage is -39v i measured. What should I do to get it to be a bit higher. maybe -20? with it at -39 i am getting 33ma on the tubes. i want 37ma for a 70% bias.

I am rather new to this and confirm if this is correct. tubewell suggests I hook a low voltage ac current into the secondary taps, 4, 8, 16 ohm, of the OT transformer and measure that from the center tap? and find the ratio between the voltage in and voltage out? i know the center tap resistance is about 3.6 which is close the original at 4.3k

http://www.classictone.net/40-18025.html - my ot

question 1: 1K so you wouldn't have to jack with the bias supply - i ran sim and show that lowest setting was about -35V, but you're close enough. if you'd like less, change the 10k under tghe bias pot to 9.1K or 8.2k. it would be best if you didn't mess with the bias supply other than make adjustments with the trimpot.

question 2: see response above.


question 3: you're go to go with that classictone OT as is. was assuming it was the original OT, might have had a higher reflected impedance than 4.3k. i assume you have it wired in correctly, if it squeals, then swap the plate leads.


to answer your third question, you'd use some clip leads to tack in a 12.6V or 6.3V transformer winding from pin 3 of one output socket tube to pin 3 of the other output tube socket. measure the AC voltage from Pin 3 to pin 3 then measure ACV from each output tap to ground with no load connected. also, i assume you know that the amp is t be unplugged & powered off, the filter caps bled down, and the power tubes removed before you perform the preceding. you can measure from CT to plate (pin 3) and then double the reading, however, usually the easiest is just to measure from pin 3 to pin 3. push-pull transformer impedance ratings are from plate to plate - e.g. if you have a 4K transformer each tube sees 2K to CT.


--pete

[Pochie45566]

hack up of your schematic with suggested mods attached.


--pete

you are so helpful! three questions.

Why a 1k there? would a 10k not be better

the bias pot minimum negative voltage is -39v i measured. What should I do to get it to be a bit higher. maybe -20? with it at -39 i am getting 33ma on the tubes. i want 37ma for a 70% bias.

I am rather new to this and confirm if this is correct. tubewell suggests I hook a low voltage ac current into the secondary taps, 4, 8, 16 ohm, of the OT transformer and measure that from the center tap? and find the ratio between the voltage in and voltage out? i know the center tap resistance is about 3.6 which is close the original at 4.3k

http://www.classictone.net/40-18025.html - my ot

question 1: 1K so you wouldn't have to jack with the bias supply - i ran sim and show that lowest setting was about -35V, but you're close enough. if you'd like less, change the 10k under tghe bias pot to 9.1K or 8.2k. it would be best if you didn't mess with the bias supply other than make adjustments with the trimpot.

question 2: see response above.


question 3: you're go to go with that classictone OT as is. was assuming it was the original OT, might have had a higher reflected impedance than 4.3k. i assume you have it wired in correctly, if it squeals, then swap the plate leads.


to answer your third question, you'd use some clip leads to tack in a 12.6V or 6.3V transformer winding from pin 3 of one output socket tube to pin 3 of the other output tube socket. measure the AC voltage from Pin 3 to pin 3 then measure ACV from each output tap to ground with no load connected. also, i assume you know that the amp is t be unplugged & powered off, the filter caps bled down, and the power tubes removed before you perform the preceding. you can measure from CT to plate (pin 3) and then double the reading, however, usually the easiest is just to measure from pin 3 to pin 3. push-pull transformer impedance ratings are from plate to plate - e.g. if you have a 4K transformer each tube sees 2K to CT.


--pete

Aah. Considering the bias supply (E) is before point A I dont see how changing the 20k to a 1k would impact the bias supply. Would the 1k not just impact B, C, D? But hey, I am awfully new to this and will take your word on it. I love learning about all of this. I'll make your modification of the 20k to a 1k and get back to you!

On the schematic you gave me of the laney 60 it is a 10k however. You are so helpful. I totally appreciate it. 

[DummyLoad]

if you run the screens at close to B+ then the bias range you have now should work. if you install the 10K from node A to B then the screen voltage is reduced possibly requiring you to tweak the bias supply as in you MAY need less than -39V to bias the amp. i was trying to keep the mods to a minimum. that's the reason i suggested the 1K and not the 10k for the node A to node B resistor. added bonus you can get away with using lower wattage resistor. a 1K 5W will work and run cool. with a ~500V B+ i'd run the tubes at 60% perhaps less.

--pete


P.S. the node E is a completely different supply and only shares a tap of the same secondary of the PT for source power. changes to affect nodes A - D of B+ power supply rails have NO effect on node E as it is a separate power supply.

[Pochie45566]

if you run the screens at close to B+ then the bias range you have now should work. if you install the 10K from node A to B then the screen voltage is reduced possibly requiring you to tweak the bias supply as in you MAY need less than -39V to bias the amp. i was trying to keep the mods to a minimum. that's the reason i suggested the 1K and not the 10k for the node A to node B resistor. added bonus you can get away with using lower wattage resistor. a 1K 5W will work and run cool. with a ~500V B+ i'd run the tubes at 60% perhaps less.

--pete


P.S. the node E is a completely different supply and only shares a tap of the same secondary of the PT for source power. changes to affect nodes A - D of B+ power supply rails have NO effect on node E as it is a separate power supply.

thanks! I will tell you if i need to adjust the bais range because right now the maximum range is from -70 to about -40 measuring from pin 5. Why not go for a 70% bias however? sorry for all the questions!

[PRR]

OK, 20K, I can't read.

> how do you know it was doubled?

Guess.

Working on thumbs, the four little stages probably suck 6mA and the 6CA7/EL34 screens probably suck 8mA. So more than double.

At idle, 6mA+8mA= 14mA, which in 20K is a *280V* drop!! A massive drop from ~~400V, apparently leaving 120V on screens, which sure would reduce big tube current. In fact all the tube currents will fall with voltage, sorta split the difference, end up low but not way-low.

OTOH I am not hot to know your actual load impedance. Between straight pentode and Ultralinear there is not a large change in optimum loading. The UL OT should be fine, screen taps taped off.

Are there any voltage measurements here or only currents?

[Pochie45566]

OK, 20K, I can't read.

> how do you know it was doubled?

Guess.

Working on thumbs, the four little stages probably suck 6mA and the 6CA7/EL34 screens probably suck 8mA. So more than double.

At idle, 6mA+8mA= 14mA, which in 20K is a *280V* drop!! A massive drop from ~~400V, apparently leaving 120V on screens, which sure would reduce big tube current. In fact all the tube currents will fall with voltage, sorta split the difference, end up low but not way-low.

OTOH I am not hot to know your actual load impedance. Between straight pentode and Ultralinear there is not a large change in optimum loading. The UL OT should be fine, screen taps taped off.

Are there any voltage measurements here or only currents?

considering this thing has been modded to hell and back, I dont care about keeping originality. Im gonna either put a 10k or a 1k there and adjust the 10k resistor before the bias control maybe to a 22k or so i have more range because right now i am maxed out on the bias trimmer

[DummyLoad]

i'd HIGHLY recommend that you DO NOT put a 10K between NODE A and NODE B. you'll need at least a 15-20W resistor if you make it a 10K.


you say you need more range on the 10k bias TRIMMER: if you have -39V to -46V now with 10k under the trimmer, and you want the EL34 to draw more current, then you need LESS than -39V so you make the 10k under the trimmer pot SMALLER AND NOT LARGER. making the 10k larger will move the range to -fourtysomething to -fiftysomething or possibly even higher.


at this point i have no more advice to offer.


--pete

[tubeswell]

If you want more negative bias voltage, then wire the bias supply as a (negative) voltage doubler.