Hi, New here and new to tubes. I have a degree in electronics, They didn't teach tubes anymore when I went
Here is my problem. My phase splitter seems to be gating(or something). It will not produce output with less than a 18V P-P input. Sounds great with the preamp cranked full power. The problem is trying to play clean or even softly(By the way home made guitar amp). I will try to post a sad looking schematic of it. If someone could take a look at it and tell me what you see I would be very greatful. Any comments are welcome on any part of the design.
Here is my problem. My phase splitter seems to be gating(or something). It will not produce output with less than a 18V P-P input. Sounds great with the preamp cranked full power. The problem is trying to play clean or even softly(By the way home made guitar amp). I will try to post a sad looking schematic of it. If someone could take a look at it and tell me what you see I would be very greatful. Any comments are welcome on any part of the design.Attachments
You've got 212 volts on the top plate of the phase splitter and 135 volts on the bottom plate. Something is wrong here. These should be within 10%. My gut feeling is you're not pulling enough current through the phase splitter. I would go with a lower value tail resistor (the 47k).
See the Fender 5F6A bassman for a nice sounding phase splitter for 6L6.
http://www1.korksoft.com/~schem/fenderamps/bassman_5f6a_schem.pdf
The Fender AB763 phase splitter is good too.
http://www1.korksoft.com/~schem/fenderamps/pro_ab763_schem.pdf
6L6 screen resistor:
100k seems kind of high here. Most common is a separate 470 or 1k resistor to each screen. I use 1k / 5watt resistors here.
2nd and 3rd gain stage:
I noticed the 250k plate resistors. Do you really need all that gain? I would reduce the plate resistors to 100k. Use separate 1.5k cathode resistors on the 2nd and 3rd stage. You might want to leave the 3rd stage without a cathode bypass cap. Keep the 470k grid stopper resistors, that's a good idea to prevent blocking distortion.
I rarely use more than a 4.7uf cathode bypass cap in guitar amps. Often, I use 1.5k / 1uf or 2.7k / .68uf. This will cut bass response and make a tighter sounding amp.
I would move the master volume control to after the 3rd stage, before the phase splitter. this would give you the following topology:
gain stage -> gain control-> gain stage -> tone control -> gain stage -> master volume -> phase splitter -> power tubes.
You should be able to get a good clean sound out of this as well as a high gain sound, and everything in between.
Tone control:
You have a simple high cut tone control here. You already have a high cut control on your guitar. I prefer the more usual bass, treble, mid controls, myself. You may be limited by available panel space. Here is a site with a good variety of tone controls
http://amps.zugster.net/articles/tonestacks/
Gain / master volume pots: these are 500k pots with 470k in parallel. I'm not sure what you're trying to accomplish here. Most guitar amps use 1Meg audio taper pots without parallel resistors.
Hope this helps,
Mike D.
See the Fender 5F6A bassman for a nice sounding phase splitter for 6L6.
http://www1.korksoft.com/~schem/fenderamps/bassman_5f6a_schem.pdf
The Fender AB763 phase splitter is good too.
http://www1.korksoft.com/~schem/fenderamps/pro_ab763_schem.pdf
6L6 screen resistor:
100k seems kind of high here. Most common is a separate 470 or 1k resistor to each screen. I use 1k / 5watt resistors here.
2nd and 3rd gain stage:
I noticed the 250k plate resistors. Do you really need all that gain? I would reduce the plate resistors to 100k. Use separate 1.5k cathode resistors on the 2nd and 3rd stage. You might want to leave the 3rd stage without a cathode bypass cap. Keep the 470k grid stopper resistors, that's a good idea to prevent blocking distortion.
I rarely use more than a 4.7uf cathode bypass cap in guitar amps. Often, I use 1.5k / 1uf or 2.7k / .68uf. This will cut bass response and make a tighter sounding amp.
I would move the master volume control to after the 3rd stage, before the phase splitter. this would give you the following topology:
gain stage -> gain control-> gain stage -> tone control -> gain stage -> master volume -> phase splitter -> power tubes.
You should be able to get a good clean sound out of this as well as a high gain sound, and everything in between.
Tone control:
You have a simple high cut tone control here. You already have a high cut control on your guitar. I prefer the more usual bass, treble, mid controls, myself. You may be limited by available panel space. Here is a site with a good variety of tone controls
http://amps.zugster.net/articles/tonestacks/
Gain / master volume pots: these are 500k pots with 470k in parallel. I'm not sure what you're trying to accomplish here. Most guitar amps use 1Meg audio taper pots without parallel resistors.
Hope this helps,
Mike D.
Yes, you want to run around a milliamp per section for that phase splitter. The plate voltages should end up about 15-20V apart. For a guitar amp, I'd probably not worry too much about AC balance and use the same value resistor on each plate. Also, you don't indicate any grid leak resistance between the o/p tube grids and the bias pot. Hopefully, that was just something omitted from the drawing.
What you're doing with the feedback is confusing me- why are you running it back to a common-mode point?
What you're doing with the feedback is confusing me- why are you running it back to a common-mode point?
Thank you
This is my first amp(besides modifications of others). I will try the things you suggest. yes the tone control is due to the amount of pots, I started with a McMartin PA amp. That tone circuit is just a slap job for testing, I will improve it when a get a little further. The 470Ks across the pots are to keep it from motor boating.
Yes the gid leak resistors are omitted
The feed back is modeled after Marshall/Matchless.
What should the output transformer ohm out to for these tubes(primary and secondary)?
Thank you again for your help, Nobody I know knows tubes so it means alot to me.
Tim
This is my first amp(besides modifications of others). I will try the things you suggest. yes the tone control is due to the amount of pots, I started with a McMartin PA amp. That tone circuit is just a slap job for testing, I will improve it when a get a little further. The 470Ks across the pots are to keep it from motor boating.
Yes the gid leak resistors are omitted
The feed back is modeled after Marshall/Matchless.
What should the output transformer ohm out to for these tubes(primary and secondary)?
Thank you again for your help, Nobody I know knows tubes so it means alot to me.
Tim
OK, rewired and can now play clean and at low levels(Thank you). Cranked up it sounds real bad(as in not good kind of bad). This is why I ask about the output transformer(As the waves going into the power tubes look good on the scope). It measures nowhere near that high(around 450 Ohms). As I said it was an old PA. It said it was 35 Watts. The way it is wired now should be around 50 Watts,Yes? I started to think about this and I think I am getting primary coil saturation or have a bad transformer. Does this sound right to you? Is a new transformer in order?
Tim,
I also have a McMartin PA, model MA-50. I bought it at a hamfest for $25. I converted it into a bass amp. The one I have runs the power tubes at a savage voltage of about 625v. It used an OD3 gas tube to supply a screen voltage of about 400v. I ditched the noisy OD3 and made a semi-regulated screen supply with a couple of TIP50 transistors. I run el-cheapo Sovtek 5881s in there with no problems in 4 years of weekly use.
Anyway getting back to your transformer, you can do a poor man's impedence measurement with a variac and a voltmeter.
see http://www.gbronline.com/radioguy/outimp.htm
Make sure you are using the correct secondary tap. PA amps have a 70v secondary tap for distributed speaker systems and this tap does not sound too stellar for guitar speakers.
Assuming the transfomer is ok and it still sounds bad cranked, it could be crossover distortion. Your schematic shows 450v on the 6L6 plates. Set the bias so that each tube is pulling about 30-35ma. This biases your tubes at about 50% of their max plate dissipation of 30 watts. That's a good starting place for bias in a guitar amp. The PA amp is normally biased much closer to class B and does sound bad cranked.
OK, we're using the correct secondary tap, bias is set, does it still sound bad? try disconnecting the NFB. Still sound bad? maybe some oscillations. Try a 47pf cap accross the phase splitter plates like the 5F6A Bassman.
Still bad? Well it could be a bad transformer. I've seen it happen. I fixed a gibson stereo guitar amp where one channel sounded like poop and the other was marvelous. I really beat my head against the wall until I swapped the transformers and discovered the problem was indeed the transformer.
Regards,
Mike D.
I also have a McMartin PA, model MA-50. I bought it at a hamfest for $25. I converted it into a bass amp. The one I have runs the power tubes at a savage voltage of about 625v. It used an OD3 gas tube to supply a screen voltage of about 400v. I ditched the noisy OD3 and made a semi-regulated screen supply with a couple of TIP50 transistors. I run el-cheapo Sovtek 5881s in there with no problems in 4 years of weekly use.
Anyway getting back to your transformer, you can do a poor man's impedence measurement with a variac and a voltmeter.
see http://www.gbronline.com/radioguy/outimp.htm
Make sure you are using the correct secondary tap. PA amps have a 70v secondary tap for distributed speaker systems and this tap does not sound too stellar for guitar speakers.
Assuming the transfomer is ok and it still sounds bad cranked, it could be crossover distortion. Your schematic shows 450v on the 6L6 plates. Set the bias so that each tube is pulling about 30-35ma. This biases your tubes at about 50% of their max plate dissipation of 30 watts. That's a good starting place for bias in a guitar amp. The PA amp is normally biased much closer to class B and does sound bad cranked.
OK, we're using the correct secondary tap, bias is set, does it still sound bad? try disconnecting the NFB. Still sound bad? maybe some oscillations. Try a 47pf cap accross the phase splitter plates like the 5F6A Bassman.
Still bad? Well it could be a bad transformer. I've seen it happen. I fixed a gibson stereo guitar amp where one channel sounded like poop and the other was marvelous. I really beat my head against the wall until I swapped the transformers and discovered the problem was indeed the transformer.
Regards,
Mike D.
I was already using the 8ohm tap, had bias set to 35ma ect.. I was sitting there looking at it (you know, open the hood and stare at it. Maybe it will work ) When I saw two things. One: I had 150K instead of 1.5K going to ground on the 12ax7
Two: Right beside me was my Peavey 260 solid state power amp, I found a way to test the preamp sections away from the power tubes. I started going from one gain stage to the next into the Power amp. Found the problem in stage 3. The only way I found to fix it was to insert a 1Mohm resistor from the screen to ground(A trick I copied from my Peavey Bravo 112 tube amp. I know, to much Peavey stuff), cleans right up. Now if I can just get rid of the noise at idle. The power section is not bad, the pre section is very noisey. The 6av6 is very microphonic, Is this normal for this tube or should I look for a new one? I may have a mullard one of these around Thanks again for your help.
) When I saw two things. One: I had 150K instead of 1.5K going to ground on the 12ax7 Two: Right beside me was my Peavey 260 solid state power amp, I found a way to test the preamp sections away from the power tubes. I started going from one gain stage to the next into the Power amp. Found the problem in stage 3. The only way I found to fix it was to insert a 1Mohm resistor from the screen to ground(A trick I copied from my Peavey Bravo 112 tube amp. I know, to much Peavey stuff), cleans right up. Now if I can just get rid of the noise at idle. The power section is not bad, the pre section is very noisey. The 6av6 is very microphonic, Is this normal for this tube or should I look for a new one? I may have a mullard one of these around Thanks again for your help.I had 150K instead of 1.5K going to ground on the 12ax7
Yep, that would account for a pretty ugly sound.
Found the problem in stage 3. The only way I found to fix it was to insert a 1Mohm resistor from the screen to ground(A trick I copied from my Peavey Bravo 112 tube amp.
I'm confused here, is stage 3 the second half of the 12AX7 preamp? Did you mean to say grid instead of screen?
The 6av6 is very microphonic, Is this normal for this tube or should I look for a new one?
This tube was common as detector / AVC in the "All American 5" type AM radio. This is the final stage before the power tube. While the triode curves are similar to 12AX7, I don't believe the heater winding is specifically wound for low hum operation like the 7025/12AX7. Thus, this tube may not be suitable as the first gain stage of a high gain preamp like yours. You may find a good one if you sort through 10 of them. Or maybe not.
I would personally not use the 6AV6. Ideally you'd put a 9 pin socket in there and use a good old 12A_7. If you must keep the 7 pin socket you could use a 6AU6 pentode wired as a triode. This tube tends to be microphonic too, but not too bad wired as a triode.
Another option for the first gain stage is the FET/MOSFET cascode.
see http://www.blueguitar.org/new/articles/other/hi_v_ss.pdf
This is considered heresy among tube afficianados, but works extremely well if you want a simple, low distortion, low noise, non-microphonic gain stage that requires no filament current and easily integrates into the high voltage world of tubes.
The power section is not bad, the pre section is very noisey.
What kind of noise are we talking about here? 60hz hum? 120hz buzz? Hiss?
Regards,
Mike D.
Yep, that would account for a pretty ugly sound.
Found the problem in stage 3. The only way I found to fix it was to insert a 1Mohm resistor from the screen to ground(A trick I copied from my Peavey Bravo 112 tube amp.
I'm confused here, is stage 3 the second half of the 12AX7 preamp? Did you mean to say grid instead of screen?
The 6av6 is very microphonic, Is this normal for this tube or should I look for a new one?
This tube was common as detector / AVC in the "All American 5" type AM radio. This is the final stage before the power tube. While the triode curves are similar to 12AX7, I don't believe the heater winding is specifically wound for low hum operation like the 7025/12AX7. Thus, this tube may not be suitable as the first gain stage of a high gain preamp like yours. You may find a good one if you sort through 10 of them. Or maybe not.
I would personally not use the 6AV6. Ideally you'd put a 9 pin socket in there and use a good old 12A_7. If you must keep the 7 pin socket you could use a 6AU6 pentode wired as a triode. This tube tends to be microphonic too, but not too bad wired as a triode.
Another option for the first gain stage is the FET/MOSFET cascode.
see http://www.blueguitar.org/new/articles/other/hi_v_ss.pdf
This is considered heresy among tube afficianados, but works extremely well if you want a simple, low distortion, low noise, non-microphonic gain stage that requires no filament current and easily integrates into the high voltage world of tubes.
The power section is not bad, the pre section is very noisey.
What kind of noise are we talking about here? 60hz hum? 120hz buzz? Hiss?
Regards,
Mike D.
Mike, Stage three is the second half of 12ax7 and yes I meant to say grid. The noise is 60hz and hiss. It is from the 6av6. I may put it as the last stage to keep from amplifing it's noise in the latter stages. I thought about replacing it with a 12ax7 but the amp is steel. Don't know if it would be too much trouble. As far as the mosfet thing, Thanks but no thanks, I'm one of the people you mentioned
[color= blue]I thought about replacing it with a 12ax7 but the amp is steel. Don't know if it would be too much trouble. [/color]
It's not too much trouble. about 5 minutes with a rounded file. Faster with a Dremel or Rotozip with grinding stone. Faster still with a drill and a uni-bit.
[color= blue]As far as the mosfet thing, Thanks but no thanks, I'm one of the people you mentioned [/color]
That's cool, I used to feel the same way. But I have found mosfets very handy for:
1. reverb driver (driving a standard reverb transformer)
2. reverb recovery
3. tremolo oscillator
4. "cathode" follower (source follower) after 1st stage of AB763 style preamp to drive tone stack with low impedence. (secret weapon for fender amps! can change tone stack to 5F6A, Marshall Plexi, or Vox top boost style)
5. 1st gain stage when no chassis space available for extra tube. This stage is normally not overdriven, it exists mainly to get the guitar signal out of the noise floor. And even if it is overdriven it does not hard clip like an opamp. It rounds off the peaks in an asymmetrical manner. Not exactly like a tube but nowhere near as harsh as an opamp. Not too many opamps can output 150volts peak-to-peak either!
If I can use tubes, that's my first choice, if not I use mosfets.
No one I know can tell the difference in these applications.
Regards,
Mike D.
It's not too much trouble. about 5 minutes with a rounded file. Faster with a Dremel or Rotozip with grinding stone. Faster still with a drill and a uni-bit.
[color= blue]As far as the mosfet thing, Thanks but no thanks, I'm one of the people you mentioned [/color]
That's cool, I used to feel the same way. But I have found mosfets very handy for:
1. reverb driver (driving a standard reverb transformer)
2. reverb recovery
3. tremolo oscillator
4. "cathode" follower (source follower) after 1st stage of AB763 style preamp to drive tone stack with low impedence. (secret weapon for fender amps! can change tone stack to 5F6A, Marshall Plexi, or Vox top boost style)
5. 1st gain stage when no chassis space available for extra tube. This stage is normally not overdriven, it exists mainly to get the guitar signal out of the noise floor. And even if it is overdriven it does not hard clip like an opamp. It rounds off the peaks in an asymmetrical manner. Not exactly like a tube but nowhere near as harsh as an opamp. Not too many opamps can output 150volts peak-to-peak either!
If I can use tubes, that's my first choice, if not I use mosfets.
No one I know can tell the difference in these applications.
Regards,
Mike D.
Tim,
Since I don't know what your power supply topology looks like, I'll describe a few possiblities for your hum and noise problem.
Hum and noise in tube amps are frequently due to a few sources: bad filtering caps, AC heater supply, ground loops and meager B+ filtering. The last is the easiest to treat: add more capacitance to the supply going to the output transformers. Additionally, you can add a choke immediately before the first capacitor off the B+ regulator. The first possibility may mean replacing any suspiciously old caps to see if the hum disappears.
The heater supply can have its hum contribution reduced by floating it off ground up to around 30 to 40VDC, or you can switch it to a DC heater supply using a solid state bridge rectifier. Before you do anything, however, you might make sure all heater wiring (if AC) is via twisted pair that is not running parallel to any signal wires. The signal wires should be close to the chassis and the heater wires away from the chassis in mid air (some manufactures run these wires vice versa: heater wires against chassis, signal in mid air). If the two groups should cross, try to cross them at 90 degrees.
If you choose to float the heater AC, you need to make sure no heater supply is referenced to ground. Next, you need to create the reference voltage. For example, using a 220K in series with a 22K, the 220K goes to B+ (that supplies the output transformer) and the opposite side of the 22K goes to ground (and can be bypassed with a 100 uF/100V cap). The junction between the two (voltage divider) will supply about 40VDC if your B+ is 450VDC. If your heater has a center tap connect this to the 40VDC supply point. If not, then run a 100R from one heater rail to the 40VDC point and another 100R from the other heater rail to the same 40VDC point; this creates an artificial center tap and floats the heater supply off ground. Again, make sure no heater rail is left going to ground. You might also wish to run the 12AX7 in 6V heater mode, if not already: short pins 4 to 5 together and run heater to the 4,5 short and pin 9, and then run all tube heaters in parallel.
If these suggestions don't help, then you'll need to evaluate the chassis for ground loops, which can be a bit difficult to isolate.
Since I don't know what your power supply topology looks like, I'll describe a few possiblities for your hum and noise problem.
Hum and noise in tube amps are frequently due to a few sources: bad filtering caps, AC heater supply, ground loops and meager B+ filtering. The last is the easiest to treat: add more capacitance to the supply going to the output transformers. Additionally, you can add a choke immediately before the first capacitor off the B+ regulator. The first possibility may mean replacing any suspiciously old caps to see if the hum disappears.
The heater supply can have its hum contribution reduced by floating it off ground up to around 30 to 40VDC, or you can switch it to a DC heater supply using a solid state bridge rectifier. Before you do anything, however, you might make sure all heater wiring (if AC) is via twisted pair that is not running parallel to any signal wires. The signal wires should be close to the chassis and the heater wires away from the chassis in mid air (some manufactures run these wires vice versa: heater wires against chassis, signal in mid air). If the two groups should cross, try to cross them at 90 degrees.
If you choose to float the heater AC, you need to make sure no heater supply is referenced to ground. Next, you need to create the reference voltage. For example, using a 220K in series with a 22K, the 220K goes to B+ (that supplies the output transformer) and the opposite side of the 22K goes to ground (and can be bypassed with a 100 uF/100V cap). The junction between the two (voltage divider) will supply about 40VDC if your B+ is 450VDC. If your heater has a center tap connect this to the 40VDC supply point. If not, then run a 100R from one heater rail to the 40VDC point and another 100R from the other heater rail to the same 40VDC point; this creates an artificial center tap and floats the heater supply off ground. Again, make sure no heater rail is left going to ground. You might also wish to run the 12AX7 in 6V heater mode, if not already: short pins 4 to 5 together and run heater to the 4,5 short and pin 9, and then run all tube heaters in parallel.
If these suggestions don't help, then you'll need to evaluate the chassis for ground loops, which can be a bit difficult to isolate.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.