I think my HT is too high...
I've just finished building a monoblock baby huey as a breadboard that I plan on using to learn the basics, then build a pair of mono Hueys. I'm using a hammond 270EX PS transformer (275-0-275 @ 125 ma) in the original baby huey circuit, and my voltages are running somewhat high. My B+ is around 360V and the diffamp anode voltages are 220V.
The bias blocks are measuring 635mv across the 16 ohm R
Here is what my ultra-newbie brain is contemplating:
Up the 16 ohm resistors to 18 ohms or so in the bias blocks to reduce power dissipation in the EL84's to around 12W. This was a suggestion by Ian in an earlier post, but I can't see how it will help the diff amp voltage.
Rebuild the power supply as a choke PS per Ian's suggestions for the monoblock build. I have a 8H choke and the req'd caps. I'm hoping that this will reduce the B+ voltage somewhat.
My output tubes are JJ's and I think I also have a set of EI's, so I'm assuming that my B+ needs to be below 350V...My OPT's are rated at 20W.
Any suggestions as to the correct way to solve this problem are greatly appreciated.
I've just finished building a monoblock baby huey as a breadboard that I plan on using to learn the basics, then build a pair of mono Hueys. I'm using a hammond 270EX PS transformer (275-0-275 @ 125 ma) in the original baby huey circuit, and my voltages are running somewhat high. My B+ is around 360V and the diffamp anode voltages are 220V.
The bias blocks are measuring 635mv across the 16 ohm R
Here is what my ultra-newbie brain is contemplating:
Up the 16 ohm resistors to 18 ohms or so in the bias blocks to reduce power dissipation in the EL84's to around 12W. This was a suggestion by Ian in an earlier post, but I can't see how it will help the diff amp voltage.
Rebuild the power supply as a choke PS per Ian's suggestions for the monoblock build. I have a 8H choke and the req'd caps. I'm hoping that this will reduce the B+ voltage somewhat.
My output tubes are JJ's and I think I also have a set of EI's, so I'm assuming that my B+ needs to be below 350V...My OPT's are rated at 20W.
Any suggestions as to the correct way to solve this problem are greatly appreciated.
Boywonder,
Do the Cathode Bias Block resistor change from 16 to 18 Ohms and then have a listen to it. That mod will reduce your output tube combined anode and screen dissipation from
0.635/16 x 360V = 14.3 watts (102% of rated 14W)
down to
0.635/18 x 360V = 12.7 watts (91% of rated 14W)
OR even use 20 Ohms for extra tube life
0.635/20 x 360V = 11.4 Watts (81% of rated 14 W)
The amp will run quite happily at this although you will get more like 12 watts rather than 10.5 watts output.
See what you think of the sound first - if not real happy then contemplate the power supply change. If you are happy just leave it that way. EL84s are a tough tube and don't seem to be too fussed about having a bit extra voltage across them. There are a lot of EL84 amps around running 380V.
Don't mess with the diffamp. 200V there is OK and if you do change the main B+ later this will drop as B+ drops.
Cheers,
Ian
Do the Cathode Bias Block resistor change from 16 to 18 Ohms and then have a listen to it. That mod will reduce your output tube combined anode and screen dissipation from
0.635/16 x 360V = 14.3 watts (102% of rated 14W)
down to
0.635/18 x 360V = 12.7 watts (91% of rated 14W)
OR even use 20 Ohms for extra tube life
0.635/20 x 360V = 11.4 Watts (81% of rated 14 W)
The amp will run quite happily at this although you will get more like 12 watts rather than 10.5 watts output.
See what you think of the sound first - if not real happy then contemplate the power supply change. If you are happy just leave it that way. EL84s are a tough tube and don't seem to be too fussed about having a bit extra voltage across them. There are a lot of EL84 amps around running 380V.
Don't mess with the diffamp. 200V there is OK and if you do change the main B+ later this will drop as B+ drops.
Cheers,
Ian
Thanks Ian: I fired it up earlier for about 3 minutes and it plays rather well! I was just a little paranoid about melting my anodes.
I will swap the resistors out after a little listening, since I'm only running 102% or so. I am planning on building the second breadboard as the fixed bias mono version, and then modifying the original amp to be fixed bias.
Thanks again.
I will swap the resistors out after a little listening, since I'm only running 102% or so. I am planning on building the second breadboard as the fixed bias mono version, and then modifying the original amp to be fixed bias.
Thanks again.
I finally got my single channel monoblock breadboard up and running for a few hours, and it sounds great, better than I expected. This has been an extremely satisfying project.
So far, I've only let the smoke out of one component; unfortunately, it was one of my input transformers
That was before I figured out the PSUD software, another very cool learnig experience.
Actually I forgot about the smoked heater resistors when I accidentally mixed up the output and driver tubes in their sockets. (The sockets are mirror imaged when looking from the top of the amp!!
) I won't be doing that again...
Now I am planning on building the second monoblock as the fixed bias version, and then updating the first mono to fixed bias, after some listening comparisons.
Here are my next batch of beginner questions:
The ZVN0545A mosfets appear to be three-legged critters; how do I interpret the "middle arrow" connection between the source and the drain on the schematic?
Are the only parameters that are important for selecting the Green and Blue LEDs the voltage drops?
It appears that I will need another transformer to generate the -45V rail; does this have any current requirements? I am using the hammond 270EX for each monoblock for everything else.
So far, I've only let the smoke out of one component; unfortunately, it was one of my input transformers
That was before I figured out the PSUD software, another very cool learnig experience.Actually I forgot about the smoked heater resistors when I accidentally mixed up the output and driver tubes in their sockets. (The sockets are mirror imaged when looking from the top of the amp!!
) I won't be doing that again...Now I am planning on building the second monoblock as the fixed bias version, and then updating the first mono to fixed bias, after some listening comparisons.
Here are my next batch of beginner questions:
The ZVN0545A mosfets appear to be three-legged critters; how do I interpret the "middle arrow" connection between the source and the drain on the schematic?
Are the only parameters that are important for selecting the Green and Blue LEDs the voltage drops?
It appears that I will need another transformer to generate the -45V rail; does this have any current requirements? I am using the hammond 270EX for each monoblock for everything else.
The "middle arrow connection" in the mosfet symbol represents the "substrate".
In all of the mosfets you are ever likely to come across it is never run out as a separate connection but is connected to the source internally.
ASIDE: There are some very specialised mosfet types usually intended for VERY high speed analog switches where the substrate connection is run out separtately and these devices have 4 leads. They also have absolutely no anti-static protection on the gate, are supplied with a conductive rubber ring around the 4 leads which has to be left in place until after the devices have been soldered into the circuit, and die if you "look at them sideways". I actually use some of them in the day job to implement analog switches which switch in about 1 nano second.
The -45V (nominal) supply needs to be able to deliver about 6 mA.
Note that you also need a +ve low voltage supply which can be any where between about +15 and +75V but only needs to supply about 1 mA
You could have a separate tranny for these.
My suggestions:
Use a zener regulator off the existing High Voltage for the anywhere between +15V and +75V supply. Make it say +33V, you need about 100K from the high voltage rail to the zener cathode. Zener anode to 0V. Use a 1.5 Watt Zener. The resistor will dissipate nearly 1 watt. so use 2 off 47K 2W metal film resistors in series.
The approx -45 supply can be done with a separate tranny in 2 different ways. Note that you are drawing only 6 mA. 6mA x 45V is 0.3 VA so a really tiny transformer will do.
1) Use a tranny with 18-0-18 volt secondary and bridge recify across the the full secondary leaving the"0" centre tap unconnected. This is waht I did for my prototypes.
2) IF you are not using the 270EX 5V winding for a tube rectifier then:
Use a 15 volt secondary, 115V primary tranny. Wire it in reverse to the 270EX 5V winding. That is connect the 15V secondary to the 270EX 5V winding. You will then get about 36 to 38 Volts AC on the primary. Bridge rectify and filter that.
The Blue LED is the reference for the diff amp current source. Keep the leads to it short and away from any AC wiring so that it doen't pick up any noise.
Cheers,
Ian
In all of the mosfets you are ever likely to come across it is never run out as a separate connection but is connected to the source internally.
ASIDE: There are some very specialised mosfet types usually intended for VERY high speed analog switches where the substrate connection is run out separtately and these devices have 4 leads. They also have absolutely no anti-static protection on the gate, are supplied with a conductive rubber ring around the 4 leads which has to be left in place until after the devices have been soldered into the circuit, and die if you "look at them sideways". I actually use some of them in the day job to implement analog switches which switch in about 1 nano second.
The -45V (nominal) supply needs to be able to deliver about 6 mA.
Note that you also need a +ve low voltage supply which can be any where between about +15 and +75V but only needs to supply about 1 mA
You could have a separate tranny for these.
My suggestions:
Use a zener regulator off the existing High Voltage for the anywhere between +15V and +75V supply. Make it say +33V, you need about 100K from the high voltage rail to the zener cathode. Zener anode to 0V. Use a 1.5 Watt Zener. The resistor will dissipate nearly 1 watt. so use 2 off 47K 2W metal film resistors in series.
The approx -45 supply can be done with a separate tranny in 2 different ways. Note that you are drawing only 6 mA. 6mA x 45V is 0.3 VA so a really tiny transformer will do.
1) Use a tranny with 18-0-18 volt secondary and bridge recify across the the full secondary leaving the"0" centre tap unconnected. This is waht I did for my prototypes.
2) IF you are not using the 270EX 5V winding for a tube rectifier then:
Use a 15 volt secondary, 115V primary tranny. Wire it in reverse to the 270EX 5V winding. That is connect the 15V secondary to the 270EX 5V winding. You will then get about 36 to 38 Volts AC on the primary. Bridge rectify and filter that.
The Blue LED is the reference for the diff amp current source. Keep the leads to it short and away from any AC wiring so that it doen't pick up any noise.
Cheers,
Ian
snubber caps across rectifier diodes
Ian: What function do the 10nf 3000v caps provide across the rectifier diodes? Are these to tame transient spikes into the diodes?
Is ceramic the cap of choice in this application or will 2000v PP's also work?
Do all SS rectifier diodes stand to benefit from these bypass caps?
(FREDs, 5YVXXX, 1n4007, etc.)
Sorry if these are newb questions but searching and reading Morgan Jones hasn't enlightened me...
Thanks
Ian: What function do the 10nf 3000v caps provide across the rectifier diodes? Are these to tame transient spikes into the diodes?
Is ceramic the cap of choice in this application or will 2000v PP's also work?
Do all SS rectifier diodes stand to benefit from these bypass caps?
(FREDs, 5YVXXX, 1n4007, etc.)
Sorry if these are newb questions but searching and reading Morgan Jones hasn't enlightened me...
Thanks
These caps are as you suggest to tame diode switching noise.
They should be high voltage ceramic but 2kV rated is fine.
2KV rated Polypropylene would certainly work but might be a bit "over the top", save them for use in the audio path.
If using standard diodes (1N4007 etc) they are essential.
If using Ultrafast Soft Recovery types (what I mainly use) they still help.
Cheers,
Ian
They should be high voltage ceramic but 2kV rated is fine.
2KV rated Polypropylene would certainly work but might be a bit "over the top", save them for use in the audio path.
If using standard diodes (1N4007 etc) they are essential.
If using Ultrafast Soft Recovery types (what I mainly use) they still help.
Cheers,
Ian
Hi boywonder,
I'm in agreement with Ian here, he is giving you wise advice.
I only disagree in rectifiers and find that standard ones are the best. Just don't use too much capacitance. A capacitor that is high in capacitance for the job will only increase diode switching noise and ringing. Also, this will cause the transformer to run warmer in extreme cases.
Hi Ian,
I have not had any bad behavior with standard rectifier diodes. The ones made for switching supplies do cause trouble. Fast rectifiers work best with switching power supplies, the reason they were created in the first place. I accept that a soft recovery type may have less noise, but the noise levels created with standard diodes is easy to deal with. I am open to your experiences. Am I way off base here?
-Chris
I'm in agreement with Ian here, he is giving you wise advice.
I only disagree in rectifiers and find that standard ones are the best. Just don't use too much capacitance. A capacitor that is high in capacitance for the job will only increase diode switching noise and ringing. Also, this will cause the transformer to run warmer in extreme cases.
Hi Ian,
I have not had any bad behavior with standard rectifier diodes. The ones made for switching supplies do cause trouble. Fast rectifiers work best with switching power supplies, the reason they were created in the first place. I accept that a soft recovery type may have less noise, but the noise levels created with standard diodes is easy to deal with. I am open to your experiences. Am I way off base here?
-Chris
Chris,
I don't believe you are "way off base".
I have found that the caps worked well to suppress switching noise with standard diodes.
I also found that when I changed standard diodes to fast soft recovery types that I could hear the improvement - so these days I do both, use the fast soft recovery diodes and fit the suppression caps.
I admit I have'nt investigated this in depth with the oscilloscope.
Cheers,
Ian.
I don't believe you are "way off base".
I have found that the caps worked well to suppress switching noise with standard diodes.
I also found that when I changed standard diodes to fast soft recovery types that I could hear the improvement - so these days I do both, use the fast soft recovery diodes and fit the suppression caps.
I admit I have'nt investigated this in depth with the oscilloscope.
Cheers,
Ian.
Hi everyone,
I was playing with the shunt feedback resistors for a while.
The difference on the performance is quite obvious and I would like to have a changeable setup. I used a potentiometer for the global feedback at the beginning and the effect was very nice, but since I liked the sound with no gNFB best I omitted that later.
Now I am wondering if a similar setup would be possible for the shunt feedback. Obviously the power requirements make it impossible to use a potentiometer (would have to be stereo).
Ideally I would like to have a setup that I can switch while playing.
I was thinking about a 2W resistor that is paralleled with a rotary switch to dial in different parallel resistors. This way there is always at least one resistor (the one with the most feedback) in the path.
Would the sudden change in feedback level that happens when switching "live" be a problem, though?
Do I have to expect a loud "pop" and dead speakers?
Has anyone thought about that earlier or implemented something similar?
It would make the setup easier, less prone to placebo effects and I could use different feedback settings in the final version for different type of music (eventually I might reduce that to two different values - high and low feedback).
Any input is welcome!
Thanks,
Martin
I was playing with the shunt feedback resistors for a while.
The difference on the performance is quite obvious and I would like to have a changeable setup. I used a potentiometer for the global feedback at the beginning and the effect was very nice, but since I liked the sound with no gNFB best I omitted that later.
Now I am wondering if a similar setup would be possible for the shunt feedback. Obviously the power requirements make it impossible to use a potentiometer (would have to be stereo).
Ideally I would like to have a setup that I can switch while playing.
I was thinking about a 2W resistor that is paralleled with a rotary switch to dial in different parallel resistors. This way there is always at least one resistor (the one with the most feedback) in the path.
Would the sudden change in feedback level that happens when switching "live" be a problem, though?
Do I have to expect a loud "pop" and dead speakers?
Has anyone thought about that earlier or implemented something similar?
It would make the setup easier, less prone to placebo effects and I could use different feedback settings in the final version for different type of music (eventually I might reduce that to two different values - high and low feedback).
Any input is welcome!
Thanks,
Martin
bayermar said:Hi everyone,
I was playing with the shunt feedback resistors for a while.
The difference on the performance is quite obvious and I would like to have a changeable setup. I used a potentiometer for the global feedback at the beginning and the effect was very nice, but since I liked the sound with no gNFB best I omitted that later.
Now I am wondering if a similar setup would be possible for the shunt feedback. Obviously the power requirements make it impossible to use a potentiometer (would have to be stereo).
Ideally I would like to have a setup that I can switch while playing.
I was thinking about a 2W resistor that is paralleled with a rotary switch to dial in different parallel resistors. This way there is always at least one resistor (the one with the most feedback) in the path.
Would the sudden change in feedback level that happens when switching "live" be a problem, though?
Do I have to expect a loud "pop" and dead speakers?
Has anyone thought about that earlier or implemented something similar?
It would make the setup easier, less prone to placebo effects and I could use different feedback settings in the final version for different type of music (eventually I might reduce that to two different values - high and low feedback).
Any input is welcome!
Thanks,
Martin
Hi Martin,
You need to change ONLY ONE resistor, R12 in this version.
There is no DC voltage accross it (although it is at near the B+), only
AC.
If shorted: no more shunt feedback.
I suppose that an 1W potentiometer can do the job.
Yves.