Adam,
Theoretically it doesn't matter but I like to tie things down in preference to letting them float around. So a 0V connection for the OT secondary common is a good idea. If using global feedback the best place for this 0V connection is the 0V end of that 470 Ohm resistor which connects the input triode grid pin 7 to 0V (via the 1K grid stopper). So wire the secondary common to that point regardless of whether you end up using global feedback or not.
Cheers,
Ian
Theoretically it doesn't matter but I like to tie things down in preference to letting them float around. So a 0V connection for the OT secondary common is a good idea. If using global feedback the best place for this 0V connection is the 0V end of that 470 Ohm resistor which connects the input triode grid pin 7 to 0V (via the 1K grid stopper). So wire the secondary common to that point regardless of whether you end up using global feedback or not.
Cheers,
Ian
I know this isn't a troubleshooting thread, but I've been posting here, and my issue is regarding this particular circuit, so post I shall...
Problem: 450V at GZ34 pin 8.
It is built exactly as the Baby Huey schematitic dictates with the following notable changes/additions:
Hammond 372JX 300-0-300 PT 250mA rated - without resistors in AC leads, should offload about 360-370V at rectifier.
Separate 5V transformer: Hammond 166J5 1A rated
I've wired the transformer's primaries (no center tap) with the PT's primaries at the power switch.
http://www.hammondmfg.com/166.htm
Power supply is shown below.
Power at the wall is a steady 117VAC.
I really don't know how to take this one. It just doesn't seem...possible. I used the 372JX before, with a GZ34 and the shown power supply circuit. All was OK.
I've triple-checked this circuit both while building and after.
Any advice is appreciated.
Regards,
Adam
Problem: 450V at GZ34 pin 8.
It is built exactly as the Baby Huey schematitic dictates with the following notable changes/additions:
Hammond 372JX 300-0-300 PT 250mA rated - without resistors in AC leads, should offload about 360-370V at rectifier.
Separate 5V transformer: Hammond 166J5 1A rated
I've wired the transformer's primaries (no center tap) with the PT's primaries at the power switch.
http://www.hammondmfg.com/166.htm
Power supply is shown below.
Power at the wall is a steady 117VAC.
I really don't know how to take this one. It just doesn't seem...possible. I used the 372JX before, with a GZ34 and the shown power supply circuit. All was OK.
I've triple-checked this circuit both while building and after.
Any advice is appreciated.
Regards,
Adam
Attachments
I did some more testing.
With the GZ34 disconnected from the circuit, I'm reading 295V at pin 8.
I've tested the PT's secondaries and all check out.
No sound comes over either channel. No hum, no hiss.
I'll refrain from posting this issue too extensively on this thread. This is the last mention I'll make of it for now.
Thanks,
Adam
With the GZ34 disconnected from the circuit, I'm reading 295V at pin 8.
I've tested the PT's secondaries and all check out.
No sound comes over either channel. No hum, no hiss.
I'll refrain from posting this issue too extensively on this thread. This is the last mention I'll make of it for now.
Thanks,
Adam
I am definitely a newb, so don't take my advice as anything special, but shouldn't you be getting 0 volts at pin 8 with the GZ34 removed? If you are getting 295 volts you probably have made a wiring error. If it is 295 volts AC it looks like you have pin 8 connected to the high voltage secondary some how.
You did not mention, but are you using the separate 5 volt transformer for the driver CCS and the 5 volt 4amp tap off the 372JX for the filament of the GZ34?
Any photos?
Good luck!
Chris
You did not mention, but are you using the separate 5 volt transformer for the driver CCS and the 5 volt 4amp tap off the 372JX for the filament of the GZ34?
Any photos?
Good luck!
Chris
What I meant is that I was reading 295V with the circuit disconnected from the GZ34.
I figured it out. As usual, I repeatedly overlooked something simple. I had connected all EL84 bias blocks to pin 1, not pin 3.
The amp is up and singing. I'm immediately impressed! I'll post my findings and some photos here later.
Back to tweaking!
Adam
I've had the 'Huey up and running for a few hours now. I must say that this is one special amp. It has certainly met and exceeded my hopes. I am absolutely in love with its overall smoothness and listenability. The midrange is very pleasing in tone and nuance, which I think is the foundation for an emotionally involving listening experience.
I've tried two different feedback configurations:
- 22K shunt and no global
- 22K shunt and 18K global (into 8 ohm speaker load)
With global feedback removed, individual voices and acoustic instruments are more vivid and really jump out into the room but overall focus is compromised when multiple instruments are being played
With global feedback included, focus is improved but the midrange receeds - slightly but noticably. The amp also produces a mild amount of ear-burning high end hash. Would this be clipping?
These are very typical findings, I'm sure.
I'm going to order more resistors so that I can play with the feedback setting more. I'm sure there's a sweet spot to be found.
All in all, I think this is an excellent sounding circuit which deserves further attention. I'm glad I built it.
Photos to come.
Thanks again Ian for such a great circuit.
Ha det bra!
Adam
Current configuration:
-1000uF 50V Nicholson Muse KZ cathode bypass, all other small electrolytics are also Muse KZ, all are rated well over requirement
-22K Mills NIWW shunt feedback, 18K IRC metal film global feedback
-GZ34 rectified, 24uf input cap followed by a 1.5H choke into 100uf per channel
-Dynaclone Z-565 output transformers
I've tried two different feedback configurations:
- 22K shunt and no global
- 22K shunt and 18K global (into 8 ohm speaker load)
With global feedback removed, individual voices and acoustic instruments are more vivid and really jump out into the room but overall focus is compromised when multiple instruments are being played
With global feedback included, focus is improved but the midrange receeds - slightly but noticably. The amp also produces a mild amount of ear-burning high end hash. Would this be clipping?
These are very typical findings, I'm sure.
I'm going to order more resistors so that I can play with the feedback setting more. I'm sure there's a sweet spot to be found.
All in all, I think this is an excellent sounding circuit which deserves further attention. I'm glad I built it.
Photos to come.
Thanks again Ian for such a great circuit.
Ha det bra!
Adam
Current configuration:
-1000uF 50V Nicholson Muse KZ cathode bypass, all other small electrolytics are also Muse KZ, all are rated well over requirement
-22K Mills NIWW shunt feedback, 18K IRC metal film global feedback
-GZ34 rectified, 24uf input cap followed by a 1.5H choke into 100uf per channel
-Dynaclone Z-565 output transformers
LowRedMoon said:
This matches my findings concerning NFB and its effects.
No. That would be your typical pentode nastiness due to high order harmonics. This is why you need some form of NFB to tame those beasts.
Definitely do that. In fact, I include variable gNFB in my designs. I've found that the level of gNFB depends on program material.
I'll try the Zobel network, Ian.
What is interesting to me is that I only noticed the high-end hash when using global feedback.
The 'Huey is now part of a video art exhibit that my friend is holding in Philadelphia. I won't have it back for a month, so further experiments are on hold for now. I'm looking for another project in the meantime. I'd really like to try an SET.
Adam
What is interesting to me is that I only noticed the high-end hash when using global feedback.
The 'Huey is now part of a video art exhibit that my friend is holding in Philadelphia. I won't have it back for a month, so further experiments are on hold for now. I'm looking for another project in the meantime. I'd really like to try an SET.
Adam
Parts list
I made a parts list for the cheap minded people like me (~$350)! I used the above schematics, the total cost is without shipping. The parts are nothing I would call High end but they should work just fine. I wasn't sure if the type of LED was super important so I got what I thought would do the trick (it is red). The sheet has my 12b4a preamp on it as well.
dstockwell said:
dstockwell said:
I made a parts list for the cheap minded people like me (~$350)! I used the above schematics, the total cost is without shipping. The parts are nothing I would call High end but they should work just fine. I wasn't sure if the type of LED was super important so I got what I thought would do the trick (it is red). The sheet has my 12b4a preamp on it as well.
working backwards from schematic
This amp seems like it may make a great project for me to build. As long as I can get my wife's OK. I may have enough spare-parts so that I wont have to buy too much and keep costs down.
However, I have a problem in simply understanding a few of the basics, in particular how to work backwards from a schematic and part values. The Baby Huey uses B+ of 300 - 350V. Given, say B+ of 300V, how do I figure out what voltage is seen by the tube plates? How do I calculate the current? From tube datasheets, I can easily figure out what the operating voltage that for a given plate voltage and cathode current. But given simply a B+ value, how do I get these values?
The answers are probably trivially simple, involve simple Ohm's Law and datasheets. I have looked at various references such as Morgan Jones, but they tend to approach things from the bottom-up ie plate voltage and required current or operating point voltage, rather than backwards.
Thank-you,
Charlie
This amp seems like it may make a great project for me to build. As long as I can get my wife's OK. I may have enough spare-parts so that I wont have to buy too much and keep costs down.
However, I have a problem in simply understanding a few of the basics, in particular how to work backwards from a schematic and part values. The Baby Huey uses B+ of 300 - 350V. Given, say B+ of 300V, how do I figure out what voltage is seen by the tube plates? How do I calculate the current? From tube datasheets, I can easily figure out what the operating voltage that for a given plate voltage and cathode current. But given simply a B+ value, how do I get these values?
The answers are probably trivially simple, involve simple Ohm's Law and datasheets. I have looked at various references such as Morgan Jones, but they tend to approach things from the bottom-up ie plate voltage and required current or operating point voltage, rather than backwards.
Thank-you,
Charlie
Re: working backwards from schematic
It's easier to work back the other way. If you look at the bias block, which is a constant current sink, you can approximate the bias current through the EL84. The 547 transistor biases the 340. The emitter of the 547 is at 0V, so it's base must be one diode drop higher than that, or about 0.6V. Since its base is connected to the 340 emitter, that point must also be at 0.6V. So, by Ohms law, to generate 0.6V over the 16Ohm resistor will require about 38mA of current. For now, add a couple mA of current to bias the 547 and call it 40mA.
With pentodes you have an additional complication and that is the influence of the screen and its current. You'll notice from the data sheets that when the screen and plate voltage are similar, the screen current is about 10% of the plate current. So assume about 36mA of plate current and find the intersection of 36mA and 300V (or whatever B+ you want) on the plate curves. Then estimate the bias voltage from the bias curves (extrapolate between them if you need to). For a sanity check, look at the typical examples on the data sheets.
At this point your approximations will be pretty good. If you want you can take your derived values and recalculate the current through the 547 transistor (Bias voltage minus 1.2V across the 5.6K resistor) it won't be far from the original estimate). You can also subtract the bias voltage from the B+ to get actual plate voltage. Plug these new values in for bias current and plate voltage and repeat your work on the curve graph. Again, the final results won't be far from your original estimates. If the values from the second iteration are close to the first, you needn't iterate further.
Sheldon
cbutterworth said:
It's easier to work back the other way. If you look at the bias block, which is a constant current sink, you can approximate the bias current through the EL84. The 547 transistor biases the 340. The emitter of the 547 is at 0V, so it's base must be one diode drop higher than that, or about 0.6V. Since its base is connected to the 340 emitter, that point must also be at 0.6V. So, by Ohms law, to generate 0.6V over the 16Ohm resistor will require about 38mA of current. For now, add a couple mA of current to bias the 547 and call it 40mA.
With pentodes you have an additional complication and that is the influence of the screen and its current. You'll notice from the data sheets that when the screen and plate voltage are similar, the screen current is about 10% of the plate current. So assume about 36mA of plate current and find the intersection of 36mA and 300V (or whatever B+ you want) on the plate curves. Then estimate the bias voltage from the bias curves (extrapolate between them if you need to). For a sanity check, look at the typical examples on the data sheets.
At this point your approximations will be pretty good. If you want you can take your derived values and recalculate the current through the 547 transistor (Bias voltage minus 1.2V across the 5.6K resistor) it won't be far from the original estimate). You can also subtract the bias voltage from the B+ to get actual plate voltage. Plug these new values in for bias current and plate voltage and repeat your work on the curve graph. Again, the final results won't be far from your original estimates. If the values from the second iteration are close to the first, you needn't iterate further.
Sheldon
Sheldon,
I am relieved that working backwards is tricky rather than it being me missing something fundamental. I tried to do the same with the Aikido and it felt like I was trying to use one variable and a datasheet to solve two interconnected variables.
Now, I do not want Ginger to think that I am criticizing him - his design seems to have rave reviews and I would like to build it, but maybe designers should include the relevant data such as tube current, etc. when they present their schematic. This would easily allow others (ignoramuses - like me) a better understanding of their rationale. Well at least, I would be able to look at the tube datasheets and see where they got their figures - maybe the proverbial lightbulb would actually come on above my head.
Anyway, I will cost out the Baby Huey. I would like an amp for my office. I also costed El Cheapo and decided I couldn't afford it right now. I also saw the Piccolo PCL86 at Walton Audio. It looks as if it is minimal in terms of parts and can put out 4W. I'll likely not be driving very efficient speakers, although I may be able to budget some single-driver bookshelf speakers if I use cheaper drivers.
Either way, I am itching to build more tube stuff, but have to be cautious of the expense.
GINGER......How does the Baby Huey compare with an AKSA? I assume you're the "Ginger" from that forum - if not, Sorry!
Regards,
Charlie
I am relieved that working backwards is tricky rather than it being me missing something fundamental. I tried to do the same with the Aikido and it felt like I was trying to use one variable and a datasheet to solve two interconnected variables.
Now, I do not want Ginger to think that I am criticizing him - his design seems to have rave reviews and I would like to build it, but maybe designers should include the relevant data such as tube current, etc. when they present their schematic. This would easily allow others (ignoramuses - like me) a better understanding of their rationale. Well at least, I would be able to look at the tube datasheets and see where they got their figures - maybe the proverbial lightbulb would actually come on above my head.
Anyway, I will cost out the Baby Huey. I would like an amp for my office. I also costed El Cheapo and decided I couldn't afford it right now. I also saw the Piccolo PCL86 at Walton Audio. It looks as if it is minimal in terms of parts and can put out 4W. I'll likely not be driving very efficient speakers, although I may be able to budget some single-driver bookshelf speakers if I use cheaper drivers.
Either way, I am itching to build more tube stuff, but have to be cautious of the expense.
GINGER......How does the Baby Huey compare with an AKSA? I assume you're the "Ginger" from that forum - if not, Sorry!
Regards,
Charlie
Charlie,
Yes I'm the "Ginger" from the Audiocircle AKSA Forum.
Power Amp listening is really subjective particularly when we start comparing Solid State Power Amps like the AKSA and Tube Power Amps.
I have 2 of Hugh Dean's AKSA Power Amps - the AKSA55 N+ (with a couple of extra mods which were covered in the Audiocircle forum) and also the Lifeforce 55 (the first issue of the Lifeforce).
Once in a while I spend an evening cycling power amps (as a sanity check) between the 2 solid state AKSA amps, a 6V6 tube rectified version and an EL84 SS rectified Baby Huey and an 845 SET. I use a Hugh Dean GK1 preamp with a few mods/upgrades (also covered in teh Audiocircle Forum) the most notable of which (for this discussion) is to switch its output from directly off the volume control (tube buffer bypassed) or from the tube buffer output.
With the Lifeforce 55 I use the tube buffered output of the GK-1 preamp (to "warm up" the sound a bit). With the 55N+ I generally use the direct output from the volume control (as the 55N= has a bit of tube like warmth of its own) - ditto for all the tube power amps.
For shear bottom end grunt the AKSA Amps win (as you would expect as they deliver >8 times the power into 4 Ohms). The EL84 Baby Huey is not far behind despite its 10W power rating. At the end of each of these sessions I have always found myself back on the EL84 Baby Huey. Then again if I'm listening to Dylan or JJ Cale or just about any Blues stuff I go for the 6V6 tube rectified (an choke input supply) Baby Huey, its just super relaxed and easy. There is just something about the tube amps which suits my taste - mind you I will never part with the AKSAs. They are my "reference" amplifiers. Once I get a tube amp sounding (to me) better than the AKSAs I know I'm just about done.
Cheers,
Ian
Yes I'm the "Ginger" from the Audiocircle AKSA Forum.
Power Amp listening is really subjective particularly when we start comparing Solid State Power Amps like the AKSA and Tube Power Amps.
I have 2 of Hugh Dean's AKSA Power Amps - the AKSA55 N+ (with a couple of extra mods which were covered in the Audiocircle forum) and also the Lifeforce 55 (the first issue of the Lifeforce).
Once in a while I spend an evening cycling power amps (as a sanity check) between the 2 solid state AKSA amps, a 6V6 tube rectified version and an EL84 SS rectified Baby Huey and an 845 SET. I use a Hugh Dean GK1 preamp with a few mods/upgrades (also covered in teh Audiocircle Forum) the most notable of which (for this discussion) is to switch its output from directly off the volume control (tube buffer bypassed) or from the tube buffer output.
With the Lifeforce 55 I use the tube buffered output of the GK-1 preamp (to "warm up" the sound a bit). With the 55N+ I generally use the direct output from the volume control (as the 55N= has a bit of tube like warmth of its own) - ditto for all the tube power amps.
For shear bottom end grunt the AKSA Amps win (as you would expect as they deliver >8 times the power into 4 Ohms). The EL84 Baby Huey is not far behind despite its 10W power rating. At the end of each of these sessions I have always found myself back on the EL84 Baby Huey. Then again if I'm listening to Dylan or JJ Cale or just about any Blues stuff I go for the 6V6 tube rectified (an choke input supply) Baby Huey, its just super relaxed and easy. There is just something about the tube amps which suits my taste - mind you I will never part with the AKSAs. They are my "reference" amplifiers. Once I get a tube amp sounding (to me) better than the AKSAs I know I'm just about done.
Cheers,
Ian
Ginger,
Good, I thought that Gingertube and Ginger were one and the same. I love my AKSA 55N (nonplussed) and do want to upgrade to the LifeForce someday. For this reason, I have been reluctant to try anoy power amp other than the AKSA becasause I feel a strong sense of loyality to Hugh. His customer service is also the best I have found.
Now for some music - New Age / electronica (Klaus Schulze, etc.) I expect that the AKSA will outperform most other amps that I could afford to own. However, some music (folk such as Dylan, Scottish such as Capercaile, etc.) while they sound good on the AKSA may sound better with a more intimate amp. Hence my interest in a tube amp - although I have no space for one in my main system, but I can use the idea of an office amp to allow me justification to build one.
Ginger, please read Sheldon's last post to this thread and see if he's correct in his method of estimating current draw. If I build the Baby Huey, I would really like to be able to use tube rectification, unless the cheapness of solid-state recitifcation is a big difference.
Athos, I would really appreciate your Excel file. Maybe we could begin a sticky thread containing people's parts lists.
Thanks,
Charlie
Good, I thought that Gingertube and Ginger were one and the same. I love my AKSA 55N (nonplussed) and do want to upgrade to the LifeForce someday. For this reason, I have been reluctant to try anoy power amp other than the AKSA becasause I feel a strong sense of loyality to Hugh. His customer service is also the best I have found.
Now for some music - New Age / electronica (Klaus Schulze, etc.) I expect that the AKSA will outperform most other amps that I could afford to own. However, some music (folk such as Dylan, Scottish such as Capercaile, etc.) while they sound good on the AKSA may sound better with a more intimate amp. Hence my interest in a tube amp - although I have no space for one in my main system, but I can use the idea of an office amp to allow me justification to build one.
Ginger, please read Sheldon's last post to this thread and see if he's correct in his method of estimating current draw. If I build the Baby Huey, I would really like to be able to use tube rectification, unless the cheapness of solid-state recitifcation is a big difference.
Athos, I would really appreciate your Excel file. Maybe we could begin a sticky thread containing people's parts lists.
Thanks,
Charlie