Amp has no headroom and low volume

[dearborndave]

I've been working on an old (1953) Masco ME36 PA amp. I replaced the inputs and outputs with 1/4" jacks, replaced all the capacitors, except for 3 ceramic caps, and a bunch of resistors that had drifted way off their original values. The only mod I did was to remove the phono channel and use the pot for a pre-phase inverter master volume. The problem I'm having is the amp has very little headroom, it starts distorting right away and it has very little output volume for a twin 6L6 amp. Shoot, I think my blackface Champ is louder! I've checked all the tubes, even replaced the power tubes, and nothing changes. I've checked all the voltages listed on the schematic and they all check out. I even isolated the output jacks thinking that might be a problem (it didn't make a difference). I'm new to rejuvenating old amps, having built several amp kits, and doing minor work on existing amps, but I have no idea what might be wrong with this amp. Does anyone out there have any suggestions? I've attached a copy of the schematic, with the only mods being the deletion of the 470 ohm grid leak resistor (R24) before the phase inverter and replaced it with a 500K-A potentiometer for the master volume, and the inputs and outputs changed to 1/4" jacks. Oh, this problem existed before the master volume mod. Any help would be greatly appreciated! Thanks in advance.

[shooter]

how old are the tubes?
how close do your volts match up with schematic volts?
got a scope?
eliminate the inputs and outputs, use a sig-gen with a known signal amplitude, use a known good speaker.

[sluckey]

The only mod I did was to remove the phono channel and use the pot for a pre-phase inverter master volume.

Wish you had left that in. This "may" just be a case of too much preamp gain. Try connecting one of the inputs directly to it's volume control, bypassing the mic preamp. Any better?

[PRR]

how close do your volts match up with schematic volts?

"I've checked all the voltages listed on the schematic and they all check out."

The gain structure is a little extreme. Note four gain stages from input to power tube grid. We can do a guitar amp in two gain stages. Add a stage to cover mix and tone loss. This 4-stage is still too much gain. Something is overloading before you get all your precious 25 Watts.

The inputs are NOT designed for modern guitars. That grid-bias scheme is OK when signals are under 50mV-100mV, as they would be in older PA. With higher supply voltage, Fender did a few like this in the 1950s, but very soon changed to Cathode Bias for more input headroom. Lift a cathode and put 2k to 5k resistor, no bypass cap. Is that more playable?

The second stage V2 also has low headroom. However if you go into the Phono input it should be playable with modern guitars; with no Master Volume it should play well but with high hiss level even with Phono pot turned down.

V2 could be re-wired as Triode for less gain and more headroom.

[shooter]

I've checked all the voltages

by then all the letters started looking the same? 

[dearborndave]

Okay, first, thanks for the replies, I really appreciate it. I did what Sluckey recommended and bypassed the first gain stage and ran the guitar signal straight to the volume pot, and now I got headroom! So it would definitely seem to be a excessive gain problem. But now the volume. It is louder than it was before, but not up to a 25 watt amp volume. What do I do about the volume and how do I fix the gain problem permanently?

Lift a cathode and put 2k to 5k resistor, no bypass cap.

Okay, what tube do I lift the cathode on and put the resistor in?

The second stage V2 also has low headroom. V2 could be re-wired as Triode for less gain and more headroom.

Could I just remove V2 from the circuit? Wouldn't that eliminate the gain problem?

Wish you had left that in. This "may" just be a case of too much preamp gain. Try connecting one of the inputs directly to it's volume control, bypassing the mic preamp. Any better?

Yes, this gave me great headroom and the amp sounds good, but still low on output.

how old are the tubes?
how close do your volts match up with schematic volts?
got a scope?
eliminate the inputs and outputs, use a sig-gen with a known signal amplitude, use a known good speaker.

The tubes are original to the amp, mid 1953 dated. I've checked them with my Sencore TC154 tube checker and they all checked good. I replaced the 6L6's with brand new NOS JAN tubes and it made no difference. The voltages match within 10% of the listed values on the schematic. I do not have a scope or a signal generator, I am fairly new to this and have just basic volt ohm meters. The speaker I used is a known good speaker.

Thanks again for the help!

[shooter]

since the volts match the schemo, my math shows as you say ~20-25W audio, 45.6W idle dissipation.

so, the 25vdc at the cathode (Vk) is "how much" drive you can provide before you start compressing the PA tubes, so the question, how much drive do you have, and how distorted is it?  you might be able to get the jist measuring Vac to the left side of the caps feeding your PA tubes, so plate side of PI.  You want a GOOD meter that can measure AC that's floating on 300vdc without loading down the system.

[dearborndave]

Thanks for the reply Shooter. I did some more measuring after seeing your reply and this is what I got.

my math shows as you say ~20-25W audio, 45.6W idle dissipation.

According to the meter I have attached to my variarc, the amp is pulling 22.5 watts at idle. I don't know if this is the same as "idle dissipation" but it's what measurement the meter on my variarc supplies.

the 25vdc at the cathode (Vk) is "how much" drive you can provide before you start compressing the PA tubes

Shame on me, but I totally missed the 25 vdc value listed on the schematic for the cathode of the power tubes and never checked it, but I just checked it and it measures 21.5 vdc

you might be able to get the jist measuring Vac to the left side of the caps feeding your PA tubes, so plate side of PI.

Okay, I just checked this and I'm getting 87 vdc on the plate of the PI (schematic list 85 vdc) with 1.1 vac

Does this info help? Thanks again!

[PRR]

> what tube do I lift the cathode on and put the resistor in?

Wire the input like a Fender. 6SC7 is a funny looking 12AX7 (pinout different). I have stolen part of a Trinity Tramp and pasted onto the ME36. Not quite Fendery 100k+1k5 values but 270k+2k7 is a common alternative. Omit cathode cap until we get the overload controlled. I also show snipping V2 G2 cap and tying G2 to plate to tame that stage.

[shooter]

the 1.1vac doesn't even effect the PA tubes.  The datasheet shows at 360 plate, with a bias ~ 22 volts (in your case, self biased, that's Vk) a Peak grid to grid drive of ~35 to 40 Vac.
so, listen to PRR's suggestions to get things dialed in so you can drive the PA tubes.

[silverfox]

> what tube do I lift the cathode on and put the resistor in?

Wire the input like a Fender. 6SC7 is a funny looking 12AX7 (pinout different). I have stolen part of a Trinity Tramp and pasted onto the ME36. Not quite Fendery 100k+1k5 values but 270k+2k7 is a common alternative. Omit cathode cap until we get the overload controlled. I also show snipping V2 G2 cap and tying G2 to plate to tame that stage.

If I'm seeing this right, the screen resistors are connected unconventionally on the pentode gain stages as well as the PI. Would a Dynaco scheme work here? i.e. by pass cap and resistor to ground work here? I could be reading the schematic wrong.

[dearborndave]

Wire the input like a Fender. 6SC7 is a funny looking 12AX7 (pinout different). I have stolen part of a Trinity Tramp and pasted onto the ME36. Not quite Fendery 100k+1k5 values but 270k+2k7 is a common alternative. Omit cathode cap until we get the overload controlled. I also show snipping V2 G2 cap and tying G2 to plate to tame that stage.

Okay PRR, I already wired the input like a fender when I installed the 1/4" jacks (I took my cues from a Fender 5A3 Deluxe which uses a 6SC7 for V1). I deleted the .005 mF caps (C4 and C5) and put 68K grid stopper resistors in, but I didn't change the 5.6 Meg input resistors (R7 and R9). So I will change them to 1 Meg. And just so I got this straight, I should install a 270K resistor on the cathode of V1a (should I do this for V1b too?), tie the screen grid of V2 to the plate, and remove the V2 screen grid .1mF cap (C8)? Is this all correct? Thanks for your help

[dearborndave]

Okay, now I have more problems! I was just farting around and taking some measurements in the amp, and there is supposed to be 0 vdc to the inputs of the power tubes, instead I'm getting 27 vdc to the input of V5 and about 60 vdc (and climbing) to the input of V6!!! What is causing this? Please help!

[sluckey]

Okay, now I have more problems! I was just farting around and taking some measurements in the amp, and there is supposed to be 0 vdc to the inputs of the power tubes, instead I'm getting 27 vdc to the input of V5 and about 60 vdc (and climbing) to the input of V6!!! What is causing this? Please help!

The only logical way for that much voltage to get to pin 5 of those output tubes would be for C14 and C15 to be leaky. Disconnect the end of both those caps from pin 5 of the 6L6s and leave the end dangling in the air. Turn on the amp and measure the voltage on the dangling cap leads. If more than a couple tenths of a volt on either, then replace the caps.

But before you do this, take a close look at your meter and be absolutely sure it's not showing you 27mV and 60mV rather than 27V and 60V.

[dearborndave]

But before you do this, take a close look at your meter and be absolutely sure it's not showing you 27mV and 60mV rather than 27V and 60V.

Duh, I feel soooo stupid. Yeah the meter was reading in mV. Thanks Sluckey, I've been poring over this amp all day and I think my brain is getting a little fried!

[sluckey]

Hey, take a break. You're not on a short timeline are you? 

[dearborndave]

Hey, take a break. You're not on a short timeline are you? 

No, I quit for the night. Sometimes I just get so wrapped up in it, I don't know when to stop

[shooter]

I don't know when to stop

this is when 

Yeah the meter was reading in mV.

anybody that's made a career outta electronic, electrical service knows, That IS when to stop, cuz the next measurement might be your last.
I "unwind" my brain by just notating a "shopping" list for tomorrow.  then tomorrow, I look, shake my head, think I had one to many, and get on with fixin things 

[PRR]

> If I'm seeing this right, the screen resistors are connected unconventionally on the pentode gain stages as well as the PI.

I see V2 as an absolutely conventional pentode connection. What looks odd to you?

V3 is pentode wired triode. The 6SJ7 pentode makes a very fine audio triode and you often find this done. Call it a cleaner half-12AU7.

[shooter]

What looks odd to you?

quick hijack;

never built with pentode pre so, why 32vdc plate and 55vdc G2/  I know you only need to swing a few Vac clean.  keeps in more linear on the tube curves?

thanks
dave

[PRR]

The lower voltages make more gain. If we wanted Strong Output we would run higher voltages.

[dearborndave]

Wire the input like a Fender. 6SC7 is a funny looking 12AX7 (pinout different). I have stolen part of a Trinity Tramp and pasted onto the ME36. Not quite Fendery 100k+1k5 values but 270k+2k7 is a common alternative. Omit cathode cap until we get the overload controlled. I also show snipping V2 G2 cap and tying G2 to plate to tame that stage.

Okay, sorry for taking so long to get back to this thread but I had to pick up some resistors, and there is like only one store that I know of in my area that sells electronic components and they like to take days off. Anyways, I now have the amp wired with 68K grid stop resistors on the inputs with 1 Meg input resistors (very Fendery). As per PRR, I installed 2.7K cathode resistors on V1 (still no bypass caps), I removed the V2 G2 cap, and I tied G2 to the plate. This raised the voltage to the plate of V2 from the 32 vdc listed on the schematic to 64 vdc, and the amp sounds really nice. Good headroom, but still down on power, as I'm still only seeing 1.125 vac to the power tubes. Any ideas on what I should do next? Thanks again for all the help you guys have been giving me. I'm learning a lot from you!

[st]

Please explain to us your test setup. You mentioned not owning a signal generator, so what ac are you measuring?

[dearborndave]

Please explain to us your test setup. You mentioned not owning a signal generator, so what ac are you measuring?

I have a cheap Vichy VC99 multimeter, basically a Chinese knock off of a Fluke, and an old Radio Shack 22-163 digital multimeter. If you look at the schematic (I attached it to my original post) I am trying to measure the ac signal voltage going into the grid of the power tubes (Between the plate of V4 and the .05 mF cap C14). As suggested by Shooter, I am trying to measure the ac signal coming off the plate of the PI going into the power tubes, but I don't know if my meter is good enough to do this. Shooter stated I should be seeing 35 to 40 Vac, and I'm only seeing a little over 1 volt

[John]

If you're not putting in audio signal at the input jack, you won't get any audio signal (AC voltage) anywhere else. That 1 volt AC you're getting is just tube noise I guess? Get someone to play just plain open chords on a guitar for you while you take measurements since you don't have a sig gen.


I believe PRR said something about this is a very high gain preamp? And maybe you're actually driving a tube or power tubes into cutoff? That might be worth thinking about.

[dearborndave]

If you're not putting in audio signal at the input jack, you won't get any audio signal (AC voltage) anywhere else. That 1 volt AC you're getting is just tube noise I guess? Get someone to play just plain open chords on a guitar for you while you take measurements since you don't have a sig gen.

Thanks John, that did the trick. I thought about checking the voltages while playing my guitar but had no way to do it. Then I figured I could use some jumpers and clip the leads in place. I did that, adjusted the range on my meter, cranked everything to ten, and started playing some big ol' open chords rather aggressively, and I got voltage spikes of over 70 Vac!

[John]

the 1.1vac doesn't even effect the PA tubes.  The datasheet shows at 360 plate, with a bias ~ 22 volts (in your case, self biased, that's Vk) a Peak grid to grid drive of ~35 to 40 Vac.
so, listen to PRR's suggestions to get things dialed in so you can drive the PA tubes.

In the post right above Shooter's, PRR talks about taming the gain. IF I'm thinking about this right, (ain't no expert) your 70'ish volts is overloading the power tubes, and driving them into cutoff. IOW, they're not able to handle what you're throwing at them.

[John]

how close do your volts match up with schematic volts?

"I've checked all the voltages listed on the schematic and they all check out."

The gain structure is a little extreme. Note four gain stages from input to power tube grid. We can do a guitar amp in two gain stages. Add a stage to cover mix and tone loss. This 4-stage is still too much gain. Something is overloading before you get all your precious 25 Watts.

The inputs are NOT designed for modern guitars. That grid-bias scheme is OK when signals are under 50mV-100mV, as they would be in older PA. With higher supply voltage, Fender did a few like this in the 1950s, but very soon changed to Cathode Bias for more input headroom. Lift a cathode and put 2k to 5k resistor, no bypass cap. Is that more playable?

The second stage V2 also has low headroom. However if you go into the Phono input it should be playable with modern guitars; with no Master Volume it should play well but with high hiss level even with Phono pot turned down.

V2 could be re-wired as Triode for less gain and more headroom.

[dearborndave]

the 1.1vac doesn't even effect the PA tubes.  The datasheet shows at 360 plate, with a bias ~ 22 volts (in your case, self biased, that's Vk) a Peak grid to grid drive of ~35 to 40 Vac.
so, listen to PRR's suggestions to get things dialed in so you can drive the PA tubes.

In the post right above Shooter's, PRR talks about taming the gain. IF I'm thinking about this right, (ain't no expert) your 70'ish volts is overloading the power tubes, and driving them into cutoff. IOW, they're not able to handle what you're throwing at them.

Okay, I've tried to follow everyone's advice and this is what I've got. The inputs have been wired like a Fender with 68K grid stop resistors and 1 Meg input resistors. V1a and V1b both have 2.7K cathode resistors added. V2 has been rewired to triode operation, (exactly the same as V3, tying G2 and G3 to the plate) and I changed the plate resistor to 47K and the cathode resistor to 2.2K (also exactly the same as V3). With the channel volume and master volume cranked to 10, I'm seeing about 60 Vac to the power tube grids. I know this is considerably higher than the 35-40 Vac that Shooter said I should be seeing, but I don't know how to bring it down any lower. The amp sounds great, has good headroom and breaks up nicely when cranked. But it still doesn't seem to be loud enough for a 20'ish watt amp. Maybe I'm just crazy and expecting too much, but my Hoffman Blues Jr. sure seems a lot louder than this amp. Anyone got any more ideas?

[shooter]

The amp sounds great, has good headroom and breaks up nicely when cranked.

so you're like 97%, not bad in the world of tubes.
how many speakers have you tried?  took me 3-4 try's on one amp, then wham, smiles all around.  OT are also a big factor, NFB another, PA tubes.........

If you're still playing with it on the bench, post the PA tubes Idle VDC's (no signal).
then IF you can find a signal source, say someone playing at full roar the same note over and over and... measure the VAC at speaker.  Ideal would be ~ 150mVac rms 1khz at the input, tweak knobs for most "perceived" loudness.

[dearborndave]

so you're like 97%, not bad in the world of tubes.

Cool, 97% is a lot better than where I started!

how many speakers have you tried?  took me 3-4 try's on one amp, then wham, smiles all around.

Two so far, the stock Eminence speaker in my Blues Jr. (it was sitting right next to the bench, so I figured what the heck!), and a Celestion G12T-100. I've got about a half dozen other speakers I can try so I will give them a shot.

OT are also a big factor, NFB another, PA tubes.........

I swapped out the PA tubes already, but that was before I got the gain problem under control, so I will try that again. How can I go about testing the OT? And the NFB?

If you're still playing with it on the bench, post the PA tubes Idle VDC's (no signal).

I'm getting 387 Vdc on the plates, 323 Vdc on the screen grids (G2), and 21.6 Vdc on the cathodes.

then IF you can find a signal source, say someone playing at full roar the same note over and over and... measure the VAC at speaker.  Ideal would be ~ 150mVac rms 1khz at the input.

I ordered a signal generator, so I will post the results of that test as soon as I receive it.

Tweak knobs for most "perceived" loudness.

I hate to sound stupid, but what knobs should I tweak?

Thanks again for all the help everyone has been giving me. I wouldn't be able to do this without you guys!

[John]

I'm getting 387 Vdc on the plates, 323 Vdc on the screen grids (G2), and 21.6 Vdc on the cathodes.

I think I see an issue. Your screen voltage should be about the same as your plates at idle. Try raising R33 to something like 200K instead of 20K. I've not seen a low value resistor (bleed resistor?) in a power supply, but then I'm not too experienced either. It's just that in guitar amps for max output, the screen voltage on your power tubes is pretty close to the plate at idle. At least what I'm familiar with anyway.


As a side note, some guys deliberately lower the screen voltage  so they do get an over driven tone at lower volume.

[dearborndave]

I think I see an issue. Your screen voltage should be about the same as your plates at idle. Try raising R33 to something like 200K instead of 20K. I've not seen a low value resistor (bleed resistor?) in a power supply, but then I'm not too experienced either. It's just that in guitar amps for max output, the screen voltage on your power tubes is pretty close to the plate at idle. At least what I'm familiar with anyway.

Thanks for the reply John, I really appreciate it. Replacing R33 is no big deal, but that resistor in this amp is a big honking wire wound resistor about 2" long, and 1/2" in diameter and it's bolted directly to the amp's chassis. Plus, it's rated at 20 watts! I'm pretty sure the 20 watt rating is overkill as I checked out some Fender dual 6L6 schematics and they're using two 1 watt resistors in series in their power supplies. I can probably get a 2 watt resistor or maybe a 5 watt, will that do the trick? Or am I reading too much into this and 1 watt resistors will be fine? I was going to get an assortment so I can experiment a little to get the voltages to match

[John]

No, I'd say the bigger wattage the better. They didn't waste money on components. I don't know near enough to advise you, given the size of that resistor and the fact that this is a hi-fi amp that you're turning into something else. I'm betting the everything is designed for a good reason. Maybe, lift the ground leg and tack in the biggest wattage & value resistor you got? But....


sometimes, an amp is what it is. To make it different, you almost have to start from scratch.

[sluckey]

I'm pretty sure the 20 watt rating is overkill

There's 320V across that 20K resistor. So the resistor is dissipating 5.12 watts. A safe rating would be 10 watts. 20 watts is even safer.

Try this... Temporarily disconnect that 20K resistor. Voltage will increase. I bet headroom and volume will increase also.

I have a similar situation in my RockOla conversion. My bleeder was a 5K/30W resistor. I put the resistor on a switch and called it my brown switch. Look at page 4 of this pdf...

     http://sluckeyamps.com/rocky/rocky.pdf

[dearborndave]

There's 320V across that 20K resistor. So the resistor is dissipating 5.12 watts. A safe rating would be 10 watts. 20 watts is even safer.
Try this... Temporarily disconnect that 20K resistor. Voltage will increase. I bet headroom and volume will increase also.
I have a similar situation in my RockOla conversion. My bleeder was a 5K/30W resistor. I put the resistor on a switch and called it my brown switch.

Thanks for the replies guys, I really appreciate it. Just for grins I disconnected the 20K resistor and put a 220K 2 watt resistor in it's place with some test leads. The voltage went up to just about 350 Vdc, I didn't test it with my guitar because I didn't want to run the amp for long with a 2 watt resistor in the circuit (there is in actuality 350V across that resistor). so before I turned the amp off I disconnected the test leads to put another resistor in series to raise the voltage more. Figuring if I added 200K to the circuit and it raised the voltage about 30 volts, if I added another 200K resistor, it would raise the voltage another 30 volts, and that would get me to the same voltage as there is on the plate. Well adding the extra resistor did nothing. And with no resistor in the circuit, the voltage is still 350 Vdc. After looking at your schematic Sluckey, with the switch installed, I'm thinking of just leaving the 20k resistor out of the circuit. If it's okay to switch the resistor out of the circuit, I'm guessing it's okay just to leave it out completely

[shooter]

the 200k at 400vdc is only sucking 2mA, so size isn't a big deal.  Once you're happy, I'd put it in as a bleeder.  I'm wondering if it was a way back "odd" voltage divider with the series R31 

[dearborndave]

the 200k at 400vdc is only sucking 2mA, so size isn't a big deal.  Once you're happy, I'd put it in as a bleeder.  I'm wondering if it was a way back "odd" voltage divider with the series R31 

Ok Shooter, now I'm confused. What 200K at 400Vdc are you talking about? I was talking about R33, a big wire wound 20k 20 watt resistor in the power supply. It was suggested to change this resistor to about 200K in an attempt to match the G2 voltage of the power tubes to the voltage on the plates. Well swapping in the 220K resistor did bring the voltage on G2 up from 320V to 350V (still about 30V less than the plates), but I also got the same result without any resistor at all. So after checking out Sluckey's schematic where he installed a switch to switch the bleeder in and out of the circuit, I just removed the 20K 20 watt resistor, now there is no bleeder. Do I need one or am I good without it?

[shooter]

Once you're happy, I'd put it in as a bleeder.

all I was saying use it as a bleeder, you're fine without it.  I use bleeders, I can get my hands inside in ~2min with 150k, without a bleeder I got time to make coffee, take a walk ~8-10min typical