tschrama said:
No exact schematic, but it is a variation on the El Cheapo (that thread is on the Decware forum but there is a souped up El Cheapo thread here) -- basically a well biased 12AT7 Long Tailed Pair driving a PP pair of trioded pentode power tubes. Baby Huey is a variation on this topology as well.
We substututed EL84s for the 12AQ5s, biased them into Class A and used iron and chassis from a Scott LK48. Power supply has all Solen caps (the only elco is the bypass in the output pair -- no room for the Solen we have to go there). The Scott had a negative bias that we used to get a minus supply for the LTP (one of the upgrades we have yet to do is to add a CCS). Another upgrade is to properly triode the output tubes (separting AC from DC conditions -- as in the PEARL triode trick or the variation Eddie Vaughn uses on the Karina. We'll also play with pentode operation (with regged screen supplies) and when we build one up with UL OPTs.
But even in its basic of states it is a really nice sounding amp.
dave
planet10 said:
15W is fine. the difference in SPL probably will not be noticeable. I hope to use it with a fullrange in a slim 12-14 liter floor standing cabinet (6"x6"x30") in a small 165-170 sq. ft room. I am just looking for a reliable schematic. My OPTs do NOT have UL capability though. I think I would try triode mode first and settle for 7-8W instead and only consider pentode if the SPLs are not adequate.
planet10 said:
This runs counter to all previous observations too. wonder why?
interesting feedback scheme
Hi, guys!
Well, it seems that I should get back to the drawing and designing board, since people are starting to catch up on me
Jokes aside, I truly hope that my RH amps and all the fuss about them that I have (tried to and) caused some time ago on these forum pages have turned some heads and ideas in the right direction.
Keep up the good work and remember good old me
Regards to all,
Alex
Hi, guys!
Well, it seems that I should get back to the drawing and designing board, since people are starting to catch up on me
Jokes aside, I truly hope that my RH amps and all the fuss about them that I have (tried to and) caused some time ago on these forum pages have turned some heads and ideas in the right direction.
Keep up the good work and remember good old me
Regards to all,
Alex
back on deck
Hi all - back on deck after 2 weeks away on holidays.
Just read thru the last 4 pages of posts.
Diomedian, Giaime
- as ErikdeBest stated the current sources in the EL84 cathodes will keep bias currents close enough (balanced) that you can safely run Toroidal Output Trannies. Just use 1% matched resistors for those 16 Ohm current set resistors. I would be interested to hear how you get on.
This aim of this design was originally to maximise results from cheap output trannies of limited primary inductance and large'ish leakage inductance and capacitance BUT results are so good that its worth trying it with good output trannies.
I've started another conversion, for niece Kate this time. (aside: That will be 3 down and 3 to go - niece Kristy has her 6V6 Baby Huey based sound system and niece Laura got a vintage electric guitar and amp instead). For Kate a Chinese "Music Angel" EL34 push pull is being modified to a "Baby Huey"ish circuit. EL34s in pentode mode. It uses the MOSFET source followers DC coupled from 6SL7 diffamp, like the 6V6 version BUT has 450V rail to give the diffamp better headroom. The amp was also modified to add individual bias adjusts for each output tube. Power supply final caps were also increased from 47uF to 220uF. Will advise how it goes.
I really miss the 6V6 version which has been delivered but happy to get a "thank you" bottle of bubbly from the niece who said she was "wrapped" by the sound. I think that means it was "cool".
Technically it wasn't as good as the EL84 version but it had a gloriousness all of its own and in some ways and on some music I actually preferred it.
The other big plus from the holiday is that I drove over to Melbourne (1800 km there and back) and (amoungst other things) got back my 845 SET which had been on loan to a friend for the last 12 months - so I have another "Reference Amplifier" to compare stuff against. I'm listening to it this week while rebuilding the EL34 Music Angel.
Cheers,
Ian
Hi all - back on deck after 2 weeks away on holidays.
Just read thru the last 4 pages of posts.
Diomedian, Giaime
- as ErikdeBest stated the current sources in the EL84 cathodes will keep bias currents close enough (balanced) that you can safely run Toroidal Output Trannies. Just use 1% matched resistors for those 16 Ohm current set resistors. I would be interested to hear how you get on.
This aim of this design was originally to maximise results from cheap output trannies of limited primary inductance and large'ish leakage inductance and capacitance BUT results are so good that its worth trying it with good output trannies.
I've started another conversion, for niece Kate this time. (aside: That will be 3 down and 3 to go - niece Kristy has her 6V6 Baby Huey based sound system and niece Laura got a vintage electric guitar and amp instead). For Kate a Chinese "Music Angel" EL34 push pull is being modified to a "Baby Huey"ish circuit. EL34s in pentode mode. It uses the MOSFET source followers DC coupled from 6SL7 diffamp, like the 6V6 version BUT has 450V rail to give the diffamp better headroom. The amp was also modified to add individual bias adjusts for each output tube. Power supply final caps were also increased from 47uF to 220uF. Will advise how it goes.
I really miss the 6V6 version which has been delivered but happy to get a "thank you" bottle of bubbly from the niece who said she was "wrapped" by the sound. I think that means it was "cool".
Technically it wasn't as good as the EL84 version but it had a gloriousness all of its own and in some ways and on some music I actually preferred it.
The other big plus from the holiday is that I drove over to Melbourne (1800 km there and back) and (amoungst other things) got back my 845 SET which had been on loan to a friend for the last 12 months - so I have another "Reference Amplifier" to compare stuff against. I'm listening to it this week while rebuilding the EL34 Music Angel.
Cheers,
Ian
so?Cheap Output Transformers
Hi Giaime,
When I started posting about torroidial OPTs, the ones I had in mind were from this place:
http://www.engineeringvista.com/trafomatic.htm
They have (had?) an introductory offer on some "custom 10W OTPs". Pick your primary impedance with UL or not, and two output taps - US$32 or something like that. (Sorry, I hate to make this sound like an advert...) The only reservation I had about trying these was aesthetics - maybe I'll have to come up with something unique.
The "standard" cheap OTP for this project that Ian probably had in mind is the Hammond 1608 - which you can find over here at least for around US$38.
Jeremy
Hi Giaime,
When I started posting about torroidial OPTs, the ones I had in mind were from this place:
http://www.engineeringvista.com/trafomatic.htm
They have (had?) an introductory offer on some "custom 10W OTPs". Pick your primary impedance with UL or not, and two output taps - US$32 or something like that. (Sorry, I hate to make this sound like an advert...) The only reservation I had about trying these was aesthetics - maybe I'll have to come up with something unique.
The "standard" cheap OTP for this project that Ian probably had in mind is the Hammond 1608 - which you can find over here at least for around US$38.
Jeremy
Just a hint for those building the 6V6 version with the MOSFET source followers.
I have now modified a Chinese "Music Angel" EL34 Amp (Pentode Mode) to this arrangement and have been fine tuning it over the last week (Kate's Amp).
Last night I modified the power supply to the source follower by adding a 2uF Polypropylene cap across the Mallory 40uF/500V electrolytic I used to decouple this supply - what an incredible difference, the tops became uncluttered and cleaner - the dynamic range expanded and the amp picked up rythym.
I WAS having a lot of trouble deciding between various levels of shunt feedback. Now its easy.
Rule#1 - check your power supplies FIRST.
So what about Kristy's Amp which has already been delivered?? By some fluke / good management (or whatever) I actually used a 10uF polypropylene by itself for the source follower supply decoupling in that amp.
Cheers,
Ian
I have now modified a Chinese "Music Angel" EL34 Amp (Pentode Mode) to this arrangement and have been fine tuning it over the last week (Kate's Amp).
Last night I modified the power supply to the source follower by adding a 2uF Polypropylene cap across the Mallory 40uF/500V electrolytic I used to decouple this supply - what an incredible difference, the tops became uncluttered and cleaner - the dynamic range expanded and the amp picked up rythym.
I WAS having a lot of trouble deciding between various levels of shunt feedback. Now its easy.
Rule#1 - check your power supplies FIRST.
So what about Kristy's Amp which has already been delivered?? By some fluke / good management (or whatever) I actually used a 10uF polypropylene by itself for the source follower supply decoupling in that amp.
Cheers,
Ian
Help with more tuning.
I finally finished building both channels of the Baby Huey. I must say I am totally amazed by how good this amp sounds (especially female vocals and strings). I’ve done a good bit of AB comparison with my Mark Levinson 331 in the last few weeks. The Baby Huey brings a new whole level of clarity to the music. In certain busy passages, the Baby Huey renders each instrument clearly while the ML may sound confused. Instead of a blur, I can now follow each instrument. Lyrics that was not clear is now much more clear.
In my current build, I replaced the 16ohm resisters in the bias circuit with a trim pot and carefully match the bias currents. I also removed the global negative feedback. But what made the most difference for me was the removal of the 4700uF bypass caps. The improvement was nothing short of dramatic. Taking the cap out and the mid and highs become crystal clear with much more details and dimensionality. The sound stage becomes much wider yet very focused.
This might have something to do with the budget parts I am using. I won’t go all out yet. I’ve ordered some polypropylene 1uF (Wilma) and some low ESR 470uF and 1000uF (Xicor) to try.
My main reason for choosing the DIY route is to understand how each of the circuit elements and components affect the sound so I can tune the amp to the way I like it. I’d welcome any suggestions on what else I might try to get the most out of this already great amp.
I finally finished building both channels of the Baby Huey. I must say I am totally amazed by how good this amp sounds (especially female vocals and strings). I’ve done a good bit of AB comparison with my Mark Levinson 331 in the last few weeks. The Baby Huey brings a new whole level of clarity to the music. In certain busy passages, the Baby Huey renders each instrument clearly while the ML may sound confused. Instead of a blur, I can now follow each instrument. Lyrics that was not clear is now much more clear.
In my current build, I replaced the 16ohm resisters in the bias circuit with a trim pot and carefully match the bias currents. I also removed the global negative feedback. But what made the most difference for me was the removal of the 4700uF bypass caps. The improvement was nothing short of dramatic. Taking the cap out and the mid and highs become crystal clear with much more details and dimensionality. The sound stage becomes much wider yet very focused.
This might have something to do with the budget parts I am using. I won’t go all out yet. I’ve ordered some polypropylene 1uF (Wilma) and some low ESR 470uF and 1000uF (Xicor) to try.
My main reason for choosing the DIY route is to understand how each of the circuit elements and components affect the sound so I can tune the amp to the way I like it. I’d welcome any suggestions on what else I might try to get the most out of this already great amp.
More Tuning
ctaudio,
Glad its up and running and you like it.
The bias Current Source bypass caps I used was a parallel combination of:
4700uF/50 V ordinary commecrial grade electrolytic
470uF/50V Blackgate Standard
1uF Polypheylene Sulphide (PPS) - polypropylene would be good too.
I was interested to note that you think its better without the 4700uF - I'll give that a try myself tonight.
For further tuning you might like to try adjusting the value of the 16K "cross coupling" resistro in the 220K/47K shunt feedback networks. This resistor sets the amount of local shunt feedback.
I have tried values beetween 12K and 47K. 16K is about the minimum I liked, there is a "sweet spot" with values around 16K to 20K. After that you need to go up to about 39K,43K or even 47K to notice any real difference.
As you increase that resistor you are effectively applying more feedback to the EL84s AND placing more demands on the diff amp front end so sound is changing from the sound of the EL84 toward the sound of the 12AX7 diff amp as you increase that resistor.
What Output Trannies did you use?
I looked for a picture of the prototype to post but all I could find was this one. The "Baby Huey" is the middle amp on the left.
The one in front of it is the ECL86 version from which it was derived. Others are 845SET (up the top), a pair of Parallel Push Pull EL34 Triode Mode Mono-blocks and a - shock horror - Solid State AKSA 55 Nivarna Plus (which holds it own with this lot reasonably well).
Cheers,
Ian
ctaudio,
Glad its up and running and you like it.
The bias Current Source bypass caps I used was a parallel combination of:
4700uF/50 V ordinary commecrial grade electrolytic
470uF/50V Blackgate Standard
1uF Polypheylene Sulphide (PPS) - polypropylene would be good too.
I was interested to note that you think its better without the 4700uF - I'll give that a try myself tonight.
For further tuning you might like to try adjusting the value of the 16K "cross coupling" resistro in the 220K/47K shunt feedback networks. This resistor sets the amount of local shunt feedback.
I have tried values beetween 12K and 47K. 16K is about the minimum I liked, there is a "sweet spot" with values around 16K to 20K. After that you need to go up to about 39K,43K or even 47K to notice any real difference.
As you increase that resistor you are effectively applying more feedback to the EL84s AND placing more demands on the diff amp front end so sound is changing from the sound of the EL84 toward the sound of the 12AX7 diff amp as you increase that resistor.
What Output Trannies did you use?
I looked for a picture of the prototype to post but all I could find was this one. The "Baby Huey" is the middle amp on the left.
The one in front of it is the ECL86 version from which it was derived. Others are 845SET (up the top), a pair of Parallel Push Pull EL34 Triode Mode Mono-blocks and a - shock horror - Solid State AKSA 55 Nivarna Plus (which holds it own with this lot reasonably well).
Cheers,
Ian
Attachments
I did try varying the 16K shunt feedback resistor between 16K and 20K in 1K increments. I didn't notice much difference though. But that was before I removed the 4700uF. I'll try a wider resistance range later.
When I remove the 4700uF, I did notice a little less bass. But I find the increase in detail more than compensate for the lower bass. (Even my wife, who’s not into audio at all, could immediately tell the added details.) I'm getting some low ESR 1000uF to see if I can find a better balance.
One thing I found about the global feedback is that as I apply the feedback, the base gets tighter, but I loose air. For my taste, I'd prefer an airier sound.
In an attempt to just get off my butt building this amp, I bought a power transformer that gives me only 150mA. With 38mA bias per tube, the transformer gets rather hot. I’ve tried reducing the bias current to 30mA, but that doesn't produce enough power to drive my B&W 804 speakers. I have a few CD I like that are recording at a rather low level. I'm now back at 38mA per tube with a B+ of ~300V. I have a RCRC in the PSU of 330ohm/2, 220uF, 220ohm and 100uF. A good amount of power is currently wasted in the dropping resistors. I may try increasing the B+ and reducing the bias current. This is not really tuning, but to make up for a mistake in my transformer purchase. I could just buy another transformer, but I find playing with the PSU educational and a lot more fun.
I'm using Hammond 1608 as output transformers. Did you say you were going to try the 125E?
When I remove the 4700uF, I did notice a little less bass. But I find the increase in detail more than compensate for the lower bass. (Even my wife, who’s not into audio at all, could immediately tell the added details.) I'm getting some low ESR 1000uF to see if I can find a better balance.
One thing I found about the global feedback is that as I apply the feedback, the base gets tighter, but I loose air. For my taste, I'd prefer an airier sound.
In an attempt to just get off my butt building this amp, I bought a power transformer that gives me only 150mA. With 38mA bias per tube, the transformer gets rather hot. I’ve tried reducing the bias current to 30mA, but that doesn't produce enough power to drive my B&W 804 speakers. I have a few CD I like that are recording at a rather low level. I'm now back at 38mA per tube with a B+ of ~300V. I have a RCRC in the PSU of 330ohm/2, 220uF, 220ohm and 100uF. A good amount of power is currently wasted in the dropping resistors. I may try increasing the B+ and reducing the bias current. This is not really tuning, but to make up for a mistake in my transformer purchase. I could just buy another transformer, but I find playing with the PSU educational and a lot more fun.
I'm using Hammond 1608 as output transformers. Did you say you were going to try the 125E?
My prototype is currently running some 8K Raa with 40% Ultralinear tap fully potted transformers which are 1970s vintage and despite being grain oriented steel core are 3 to 4 times the physical size of the Hammond 1608.
The transformers I have on order to try are the Hammond 1650E. These are the same impedance spec and UL taps etc. as the 1608 but are rated at 15W and the core size is about the same as the usual Hammond 25W. They do seem to have a different winding structure so who knows until I try them. I thought they might give me a bit more bass or at least cleaner bass (be less susceptible to core saturation).
I used the 1608 in the 6V6 version I built for niece Kristy.
Your observation on global feedback is exactly what I have found. I have a Pentode Mode EL34 Amp which has a "Baby Huey" ish mod. I can't get enough damping factor from that mod alone and have added some global feedback - any more than about 5 or 6 dB of global feedback kills the "air" stone dead.
Cheers,
Ian
The transformers I have on order to try are the Hammond 1650E. These are the same impedance spec and UL taps etc. as the 1608 but are rated at 15W and the core size is about the same as the usual Hammond 25W. They do seem to have a different winding structure so who knows until I try them. I thought they might give me a bit more bass or at least cleaner bass (be less susceptible to core saturation).
I used the 1608 in the 6V6 version I built for niece Kristy.
Your observation on global feedback is exactly what I have found. I have a Pentode Mode EL34 Amp which has a "Baby Huey" ish mod. I can't get enough damping factor from that mod alone and have added some global feedback - any more than about 5 or 6 dB of global feedback kills the "air" stone dead.
Cheers,
Ian
Hi Gingertube
I have also made a start with your 'baby huey', but for the moment just one channel is built and it is not quiet, yet. It's a prototype channel, in which I can easily changes things - takes more time to build, but can be easiy change things afterwards. Sorry I can't post a picture.
I recognize the parallel push pull EL34 amplifier - it is a project from Menno van der Veen/Plitron, using a single ECC82 as grounded cathode + concertina phase splitter. How does the baby huey compare to this EL34 PPP? I don't know much about tubed circuits, but I never liked MvdV designs that much - that's why I am asking.
Erik
I have also made a start with your 'baby huey', but for the moment just one channel is built and it is not quiet, yet. It's a prototype channel, in which I can easily changes things - takes more time to build, but can be easiy change things afterwards. Sorry I can't post a picture.
I recognize the parallel push pull EL34 amplifier - it is a project from Menno van der Veen/Plitron, using a single ECC82 as grounded cathode + concertina phase splitter. How does the baby huey compare to this EL34 PPP? I don't know much about tubed circuits, but I never liked MvdV designs that much - that's why I am asking.
Erik
ErikdeBest,
Yep thats them.
With the 12AU7 ditched and replaced with an ECC99 and the EL34s in triode mode they are pretty good. Ultralinear Mode did'nt hack it (and was unstable unless the screen resistors were changed to 1K) and Pentode Mode was a complete waste of time. They have been sitting on the workshop shelf for the last 12 months not being used - until I get round to a redesign/rebuild.
The "EL84 "Baby Huey" is my favourite Amp and the one I use most of the time. The 845 SET is VERY MARGINALLY better at some music so long as its not turned up to loud - it just doesn't have the flexibility and all round gorgeousness across all music selections and all volume settings of the "Baby Huey" and it cost a hell of a lot more. Also the 845 SET's 880V rails probably makes it unsuitable for newbies.
The whole idea of the "Baby Huey" was something simple, elegant, cheap, suitable for newbies and above all gorgeous. I'm glad others are finding it as stunning as I do. One CAUTION I would add is that its a reasonably low damping factor amp and may not suit difficult speakers - it almost certainly won't suit electrostatics (although I haven't tried it with electrostatics).
Cheers,
Ian
Yep thats them.
With the 12AU7 ditched and replaced with an ECC99 and the EL34s in triode mode they are pretty good. Ultralinear Mode did'nt hack it (and was unstable unless the screen resistors were changed to 1K) and Pentode Mode was a complete waste of time. They have been sitting on the workshop shelf for the last 12 months not being used - until I get round to a redesign/rebuild.
The "EL84 "Baby Huey" is my favourite Amp and the one I use most of the time. The 845 SET is VERY MARGINALLY better at some music so long as its not turned up to loud - it just doesn't have the flexibility and all round gorgeousness across all music selections and all volume settings of the "Baby Huey" and it cost a hell of a lot more. Also the 845 SET's 880V rails probably makes it unsuitable for newbies.
The whole idea of the "Baby Huey" was something simple, elegant, cheap, suitable for newbies and above all gorgeous. I'm glad others are finding it as stunning as I do. One CAUTION I would add is that its a reasonably low damping factor amp and may not suit difficult speakers - it almost certainly won't suit electrostatics (although I haven't tried it with electrostatics).
Cheers,
Ian
Hi Gingertube
thanks for the reply - and I was sure you would have made some modifications on that EL34 PPP. Menno made such nice trafos, but the amps developed by him :S I can't do better, but at least I don't publish. I once asked, here at diyaudio, about opinions for his 6AS7 push pull design - reply was not very favorable...
About the baby huey. I am using a pair of fullrange in open baffle, so a low damping factor is no problem, maybe even better. Let's see.
Erik
thanks for the reply - and I was sure you would have made some modifications on that EL34 PPP. Menno made such nice trafos, but the amps developed by him :S I can't do better, but at least I don't publish. I once asked, here at diyaudio, about opinions for his 6AS7 push pull design - reply was not very favorable...
About the baby huey. I am using a pair of fullrange in open baffle, so a low damping factor is no problem, maybe even better. Let's see.
Erik
Damping Factor of Baby Huey
I did some measurements last night with 2 different shunt feedback values.
1) With 20K for the cross connect resistor
Vout Open Circuit (Voc) = 7.09 V RMS
Vout 3.5 Ohm load (VL) = 4.46 V RMS
Zout = [(Voc - VL)/VL] * 3.5 = 2.0 Ohms (4 Ohm tap) - at 5.7 Watts out
At full power (approx 7 watts with 280V rail)
Signal at 20K 220K 47K node = 40 V pk-pk
Signal at diffamp anodes = 15 V pk-pk
that is 55 V pk-pk generated across the 220K => 250uA pk-pk
With idle of 500uA each side of the diff amp thats OK ( up to 625 and down to 375 uA).
2) With 47K for the cross connect
Vout Open Circuit Voc = 4.67 V RMS(same input as above)
Vout 3.5 Ohm Laod VL = 3.38 V RMS
Zout = 1.3 Ohms (4 Ohm tap)
At full power (as above)
Signal at 47K 220K 47K node = 72 V pk-pk
Signal at diffamp anodes = 15 V pk-pk
That is 87 V pk-pk across 220K => 395 uA pk-pk (diffamp each side up to 700uA and down to 300uA).
With the 47K it was just starting to sound a little "dry" - a bit of 3rd harmonic distortion from the diffamp maybe.
Note that:
Increase the feedback set resistor by 2.35 times
Increased the feedback voltage at 220K 47K Rfeedback node by 1.8 times
Decreased Zout by 1.5 times
Decreased Gain by 1.5 times
That amounts to an extra 3.5dB of feedback
Sorry I did'nt get to try the 4700uF deletion yet to see how it sounds in my amp.
Cheers,
Ian
I did some measurements last night with 2 different shunt feedback values.
1) With 20K for the cross connect resistor
Vout Open Circuit (Voc) = 7.09 V RMS
Vout 3.5 Ohm load (VL) = 4.46 V RMS
Zout = [(Voc - VL)/VL] * 3.5 = 2.0 Ohms (4 Ohm tap) - at 5.7 Watts out
At full power (approx 7 watts with 280V rail)
Signal at 20K 220K 47K node = 40 V pk-pk
Signal at diffamp anodes = 15 V pk-pk
that is 55 V pk-pk generated across the 220K => 250uA pk-pk
With idle of 500uA each side of the diff amp thats OK ( up to 625 and down to 375 uA).
2) With 47K for the cross connect
Vout Open Circuit Voc = 4.67 V RMS(same input as above)
Vout 3.5 Ohm Laod VL = 3.38 V RMS
Zout = 1.3 Ohms (4 Ohm tap)
At full power (as above)
Signal at 47K 220K 47K node = 72 V pk-pk
Signal at diffamp anodes = 15 V pk-pk
That is 87 V pk-pk across 220K => 395 uA pk-pk (diffamp each side up to 700uA and down to 300uA).
With the 47K it was just starting to sound a little "dry" - a bit of 3rd harmonic distortion from the diffamp maybe.
Note that:
Increase the feedback set resistor by 2.35 times
Increased the feedback voltage at 220K 47K Rfeedback node by 1.8 times
Decreased Zout by 1.5 times
Decreased Gain by 1.5 times
That amounts to an extra 3.5dB of feedback
Sorry I did'nt get to try the 4700uF deletion yet to see how it sounds in my amp.
Cheers,
Ian