Next problem is the little hum
Hi all,
A big thank you to PRR for the many suggestions. I spend a few nights trying all combination and finally settling in with:
1. The connection remain the same but added a feedback to the 12ax7 from the speaker 16ohm output terminal. This controls the higher volume or bass distortion a little. I am not sure what value to use so I just try 68Kohm resistor for the feedback path and connect them to the cathode. Tell me if this is not ok.
2. Before, there is only a single resistor of 100 ohm and a single 100uF 160V capacitor connected to both the cathode of the el84. Now I seperated the cathode with each el84 is having it's own resistor of 270ohm and a capacitor of 100uF 16V ( I only have 16V in stock ).
The clipping still happen but at a more expected level( at a much louder volume).
But when I hooked up to a more sensitive speaker and now I am hearing a hum. Actually the hum did present with the small speaker too. The hum remains the same throughout the volume level.
Any suggestion to solve this problem?
Thanks all.
Hi all,
A big thank you to PRR for the many suggestions. I spend a few nights trying all combination and finally settling in with:
1. The connection remain the same but added a feedback to the 12ax7 from the speaker 16ohm output terminal. This controls the higher volume or bass distortion a little. I am not sure what value to use so I just try 68Kohm resistor for the feedback path and connect them to the cathode. Tell me if this is not ok.
2. Before, there is only a single resistor of 100 ohm and a single 100uF 160V capacitor connected to both the cathode of the el84. Now I seperated the cathode with each el84 is having it's own resistor of 270ohm and a capacitor of 100uF 16V ( I only have 16V in stock ).
The clipping still happen but at a more expected level( at a much louder volume).
But when I hooked up to a more sensitive speaker and now I am hearing a hum. Actually the hum did present with the small speaker too. The hum remains the same throughout the volume level.
Any suggestion to solve this problem?
Thanks all.
Hi,
Somehow I doubt this.
Historically the SE xformer predates the UL type connection by at least 25 years.
I've actually never seen SE xformers with an UL tap, aren't these rather recent developments?
I don't really follow the SE scene all that much, so pls. correct me if I'm wrong.
Cheers,
Somehow I doubt this.
Historically the SE xformer predates the UL type connection by at least 25 years.
I've actually never seen SE xformers with an UL tap, aren't these rather recent developments?
I don't really follow the SE scene all that much, so pls. correct me if I'm wrong.
Cheers,
fdegrove said:If the EL84s are left in penthode mode he could then use half of the 12AX7A as a phase splitter....
For inspiration:
BONAVOLTA
Cheers,
.... but then it's going to need feedback to get a sensibly low output Z and distortion.
Something triode will be better sounding and easier, but a better driver than the 12AX7 will be required, and of course the power will be a lot lower.
The only one of the Bona ccts I liked was EL84 PP # 4.
If feedback is used, consider, Dyna 6BQ5
Dyna ST35
Eico HF86
SDS Labs
Fisher SA16
Fisher SA100
I also have a copy of Gabe's Magnavox variation, and that would be my choice, but I'm not sure he'll want me to post it. If he's OK, I'll post it later.
Second choice would be the Fisher SA16 above.
Some of the above ccts are ultra-linear, but you could modify them to tetrode, to suit your transformer.
Why not use a transistor ccs LTP under the two EL84s for PP?
The bias point of the el84s is more than high enough for a bipolar ccs...
I've tried this out and it's close to using a meduim price interstage and much cheaper...
Doesn't need a big heatsink as it should only run about one Watt diss. Just bolt it to the chassis (suitably electrically insulated of course.)
ciao
James
The bias point of the el84s is more than high enough for a bipolar ccs...
I've tried this out and it's close to using a meduim price interstage and much cheaper...
Doesn't need a big heatsink as it should only run about one Watt diss. Just bolt it to the chassis (suitably electrically insulated of course.)
ciao
James
> added a feedback to the 12ax7 from the speaker 16ohm output terminal. This controls the higher volume or bass distortion a little.
OK, but now you are getting away from the "bad old sound" that this amp was made for. Keep improving it and it won't have any particular character at all. Which may be good, but fixing a bodge-job is probably more work than doing a good amp from scratch.
Instead of feedback, try the old-timer plan: wire the output tube as Triode. Resistor-load power will be down, but usable level into a speaker may be higher because a pentode freaks-out at impedance peaks. Taming a no-feedback pentode and arbitrary speaker is hard work; a triode is naturally tame and unlikely to blatt any speaker. Disconnect whatever now goes to the output tube Screen, rig a 470Ω (or 220, or 1K) resistor from Screen to Plate.
> I am not sure what value to use so I just try 68Kohm resistor for the feedback path and connect them to the cathode.
When you did that, did gain go down?
If it went down (or if you are not sure), try 33K or even 22K.
If the gain goes up, or the amp sounds "funny", when you add feedback, reverse the primary leads on the output transformer. You have no way of knowing absolute polarity, so you can't know if feedback will be negative or positive. But negative feedback will reduce gain; positive feedback will increase gain a little then increase all the amp's flaws then just squeal. So a little trial and error will tell you. And with a transformer, it is a snap to try it the other way.
If you go way too low, like 5K, you will run into subsonic and supersonic oscillation. But I think you will be unhappy with the gain before you get "too much" feedback.
I would expect 22K or 33K to be about "normal", for gain and for a reasonable amount of feedback over an unknown transformer.
> Hum...
Hmmmm..... did you ever post a clear plan for the power supply? I would start with rectifier, 47 or 100µFd first cap, 100Ω resistor, 220µFd second cap feeding the output stage, and maybe 47K and 47µFd filtering the first stage. With a hint of feedback, and the heater leads to the 12AX7 tightly twisted and laid against the chassis running away from the 12AX7 and parts, that should be quite clean. It is possible to use much less capacitance; but at today's prices I like overkill. Many people have many opinions, but I like the sound of an amp with so much PS capacitance that it keeps playing after you turn it off.
If a solid power supply does not fix the hum, you have grounding issues (which are very common in bodged tube amps). Power transformer HV CT goes straight to first filter cap. First cathode resistor goes straight to input jack and volume pot ground. Probably jack ground should be chassis ground, but if the previous builder hard-grounded the power supply common then it may be easier to insulate the input jack from chassis. Then run a heavy bare wire from input ground to B+ ground, and tap your intermediate grounds along that wire in order of signal flow. (You see why I say it may be easier to start fron scratch than to work too hard fixing a bodge-job.)
You could do push-pull with 12AX7 and another power tube. Of course check your power transformer for heat: if it is hot now, it will be too hot with another big heater load. 6V6 and 6AQ5 are low-current heaters and generally work great for small P-P amps. Push-pull should improve bass-loss due to transformer saturation; P-P pentode without feedback is not going to cure the pentode's tendency to blatt at bass resonance and other impedance lumps. Push-pull triode needs no feedback to give great sound, unless your speaker has large impedance lumps.
OK, but now you are getting away from the "bad old sound" that this amp was made for. Keep improving it and it won't have any particular character at all. Which may be good, but fixing a bodge-job is probably more work than doing a good amp from scratch.
Instead of feedback, try the old-timer plan: wire the output tube as Triode. Resistor-load power will be down, but usable level into a speaker may be higher because a pentode freaks-out at impedance peaks. Taming a no-feedback pentode and arbitrary speaker is hard work; a triode is naturally tame and unlikely to blatt any speaker. Disconnect whatever now goes to the output tube Screen, rig a 470Ω (or 220, or 1K) resistor from Screen to Plate.
> I am not sure what value to use so I just try 68Kohm resistor for the feedback path and connect them to the cathode.
When you did that, did gain go down?
If it went down (or if you are not sure), try 33K or even 22K.
If the gain goes up, or the amp sounds "funny", when you add feedback, reverse the primary leads on the output transformer. You have no way of knowing absolute polarity, so you can't know if feedback will be negative or positive. But negative feedback will reduce gain; positive feedback will increase gain a little then increase all the amp's flaws then just squeal. So a little trial and error will tell you. And with a transformer, it is a snap to try it the other way.
If you go way too low, like 5K, you will run into subsonic and supersonic oscillation. But I think you will be unhappy with the gain before you get "too much" feedback.
I would expect 22K or 33K to be about "normal", for gain and for a reasonable amount of feedback over an unknown transformer.
> Hum...
Hmmmm..... did you ever post a clear plan for the power supply? I would start with rectifier, 47 or 100µFd first cap, 100Ω resistor, 220µFd second cap feeding the output stage, and maybe 47K and 47µFd filtering the first stage. With a hint of feedback, and the heater leads to the 12AX7 tightly twisted and laid against the chassis running away from the 12AX7 and parts, that should be quite clean. It is possible to use much less capacitance; but at today's prices I like overkill. Many people have many opinions, but I like the sound of an amp with so much PS capacitance that it keeps playing after you turn it off.
If a solid power supply does not fix the hum, you have grounding issues (which are very common in bodged tube amps). Power transformer HV CT goes straight to first filter cap. First cathode resistor goes straight to input jack and volume pot ground. Probably jack ground should be chassis ground, but if the previous builder hard-grounded the power supply common then it may be easier to insulate the input jack from chassis. Then run a heavy bare wire from input ground to B+ ground, and tap your intermediate grounds along that wire in order of signal flow. (You see why I say it may be easier to start fron scratch than to work too hard fixing a bodge-job.)
You could do push-pull with 12AX7 and another power tube. Of course check your power transformer for heat: if it is hot now, it will be too hot with another big heater load. 6V6 and 6AQ5 are low-current heaters and generally work great for small P-P amps. Push-pull should improve bass-loss due to transformer saturation; P-P pentode without feedback is not going to cure the pentode's tendency to blatt at bass resonance and other impedance lumps. Push-pull triode needs no feedback to give great sound, unless your speaker has large impedance lumps.
Listening Test
Hi all,
You guys are great! There are so much suggestions that even my head needs time to digest them.
Well, I decided to put this amp on a real listening test last week. Knowing my uncle who is good at listening and care none on circuit to get his feedback on the amp sounding.
After listening for a few days, his comments on the amp was that.
1. It is so much better than he originally got it! no distortion even at a higher volume. ( I'm surprised.. well it could be that my CDP is giving a bigger swing.
2. The clarity and mid-high preformance is very good especially for saxophone! He began to have a higher respect for this amp in this regard... despite the "cheap" looking design externally.
3. It gives a warm sound after 1.5hours of running.
4. However, it lacks in the department of bass!... Well the ultimate human ear test is correlating to the technical findings... he ask me if there is anything can be done to the bass portion without affecting the mid-high department...
I think the bass will come back with PP design.. which is having more power...welll the OTP could be design for PP purpose...
at the same time.... if there is some minimal changes that I can do to modify the tone to give a higher bass... that would be great too. I have bypass on the bass and treble control for the listening test.
I have read somewhere in this forum thread about the value of the coupling capacitance will have influence on the amount of bass that you will have ( at the expense of high ). Anyone know about this post?
Well.. currently I am on national duty ( Singapore is a citizen army ) for 2 weeks.... I will work on it once I come back from this duty.... in the mean time, I would like to build up the circuit with the many generous contributors here.
Regards,
Hi all,
You guys are great! There are so much suggestions that even my head needs time to digest them.
Well, I decided to put this amp on a real listening test last week. Knowing my uncle who is good at listening and care none on circuit to get his feedback on the amp sounding.
After listening for a few days, his comments on the amp was that.
1. It is so much better than he originally got it! no distortion even at a higher volume. ( I'm surprised.. well it could be that my CDP is giving a bigger swing.
2. The clarity and mid-high preformance is very good especially for saxophone! He began to have a higher respect for this amp in this regard... despite the "cheap" looking design externally.
3. It gives a warm sound after 1.5hours of running.
4. However, it lacks in the department of bass!...
Well the ultimate human ear test is correlating to the technical findings... he ask me if there is anything can be done to the bass portion without affecting the mid-high department...I think the bass will come back with PP design.. which is having more power...welll the OTP could be design for PP purpose...
at the same time.... if there is some minimal changes that I can do to modify the tone to give a higher bass... that would be great too. I have bypass on the bass and treble control for the listening test.
I have read somewhere in this forum thread about the value of the coupling capacitance will have influence on the amount of bass that you will have ( at the expense of high ). Anyone know about this post?
Well.. currently I am on national duty ( Singapore is a citizen army ) for 2 weeks.... I will work on it once I come back from this duty.... in the mean time, I would like to build up the circuit with the many generous contributors here.
Regards,
> the value of the coupling capacitance will have influence on the amount of bass that you will have
Yes. It is very possible you have a too-small coupling cap.
Did you ever post a complete and accurate schematic of the amplifier as it is now?
> (at the expense of high)
No. Or not so you would notice.
You need a high-capacitance capacitor (relative to the resistance at that point) to pass bass. That will not hurt highs.
Except: a high-capacitance cap will usually be physically large. And that adds stray capacitance which rolls-off the highs.
It isn't a real problem. Video amplifiers and other wide-band tube designs passed 50Hz to 5MHz without any particular problem from the caps.
If a grid resistor is 1MegΩ use at least 0.01µFd cap, and a 0.05µFd cap is not too big. For a 100K grid resistor, use 0.1µFd to 0.5µFd cap.
Cathode caps are harder to calculate since the cathode impedance of the tube is often not known with precision. But modern electrolytics are so cheap, we can just super-size them. Try for 47µFd on "small tube" cathode resistors, 470µFd on large tubes like output stages.
> I am on national duty (Singapore is a citizen army) for 2 weeks....
I wish you very good luck. I hope this is just "practice", that you will march around the camp and the only danger is the mess-hall food. Some friends of mine were involved in the US Army duty in the Mid-East desert. However we feel about President Bush's actions, I am very glad when my friends return safe and sound.
Yes. It is very possible you have a too-small coupling cap.
Did you ever post a complete and accurate schematic of the amplifier as it is now?
> (at the expense of high)
No. Or not so you would notice.
You need a high-capacitance capacitor (relative to the resistance at that point) to pass bass. That will not hurt highs.
Except: a high-capacitance cap will usually be physically large. And that adds stray capacitance which rolls-off the highs.
It isn't a real problem. Video amplifiers and other wide-band tube designs passed 50Hz to 5MHz without any particular problem from the caps.
If a grid resistor is 1MegΩ use at least 0.01µFd cap, and a 0.05µFd cap is not too big. For a 100K grid resistor, use 0.1µFd to 0.5µFd cap.
Cathode caps are harder to calculate since the cathode impedance of the tube is often not known with precision. But modern electrolytics are so cheap, we can just super-size them. Try for 47µFd on "small tube" cathode resistors, 470µFd on large tubes like output stages.
> I am on national duty (Singapore is a citizen army) for 2 weeks....
I wish you very good luck. I hope this is just "practice", that you will march around the camp and the only danger is the mess-hall food. Some friends of mine were involved in the US Army duty in the Mid-East desert. However we feel about President Bush's actions, I am very glad when my friends return safe and sound.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.