• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

EL84 Amp - Baby Huey

On the end of the 220k there is 47k//8k and a voltage 17x/7 = ~2,5 x signal on g1(EL84)=a(ECC83).

The anode ECC83 sees (220k+47k//8k)/(2,5+1)=64k and the 470k (grid leak EL84) in parallel.Makes a total dynamic load of 56k.

Looking at the curves the current could go up to 0,7mA but not more to stay out of the grid current region.
Mona

Dear Mona, I'm trying to understand more about this amp and how it works. Basically to understand how this has been taken from ECL86 and adapted to EL84, in order to do the same with other tubes.
I take your post as a starting point.

In attachment the loadline for the EL84 at 315V B+ 35 mA at idle in UL with 8k Raa: Vg1 is -12,3 V and the amp is in class A until Vg1 -6 V (that's 4,9 W) and then around 14 W going up to Vg1=0. Something more going into AB2 is be allowed by the driver.

So the driver must go:

Considering 47k and 16k shunt feedback:
Total dynamic load = 56k
ECC83 swing must be 2,5 times the one needed by the grid of the EL84

For Class A:
2x [-12,3 -(-6)] x 2,5 = 32 Vpp

For Class AB1:
2x [-12,3 -(0)] x 2,5 = 62 Vpp


Considering 47k and 33k shunt feedback:
Total dynamic load = 39k
ECC83 swing must be 4,4 times the one needed by the grid of the EL84

For Class A:
2x [-12,3 -(-6)] x 4,4 = 56 Vpp

For Class AB1:
2x [-12,3 -(0)] x 4,4 = 109 Vpp

I've attached the 12AX7 loadlines with AC loadlines at 56 and 39k.
With 33k the driver is unable to accomplish his work, while with 16k it seems that the perfect balance in terms of driving is obtained between the 12ax7 and the EL84.

So the sonic results can be due to the fact that drivers and output tubes reach their limits together? Can be shunt feedback seen (also) as a way to balance driver and output tubes in order to optimize the power amp?

Can this be used as a parameter when switching to other combos of drivers and output tubes?

Thank you in advance
Roberto
 

Attachments

  • 12AX7-315V-0,7mA-264k-56&37k.jpg
    12AX7-315V-0,7mA-264k-56&37k.jpg
    321.7 KB · Views: 491
  • BH-EL84-315VB+-35mA-8kRaa.jpg
    BH-EL84-315VB+-35mA-8kRaa.jpg
    583.2 KB · Views: 465
So the driver must go:

Considering 47k and 16k shunt feedback:
Total dynamic load = 56k
ECC83 swing must be 2,5 times the one needed by the grid of the EL84

For Class A:
2x [-12,3 -(-6)] x 2,5 = 32 Vpp

For Class AB1:
2x [-12,3 -(0)] x 2,5 = 62 Vpp
Since the anode ECC83 is AC-coupled to grid of the EL84 the voltage swing doesn't change with the feedback.The load of the ECC83 changes and so do the requiered current.More feedback more current needed to get the same voltage swing.
So the sonic results can be due to the fact that drivers and output tubes reach their limits together? Can be shunt feedback seen (also) as a way to balance driver and output tubes in order to optimize the power amp?
In a way, yes.With more gain in the output stage the dynamic load on the pre get heavyer and it's gain drops.
Since the gain of the output stage isn't a constant the feedback load changes too, making the loadline of the ECC83 curved, no straigt line !
As for balancing the gains, perhaps more or less, certainly not perfect.
Can this be used as a parameter when switching to other combos of drivers and output tubes?
Perhaps, but with output tubes witch
demand a large voltage swing at the grid the driver gets a difficult task to deliver both a large voltage and much current.
Mona
 
Since the anode ECC83 is AC-coupled to grid of the EL84 the voltage swing doesn't change with the feedback.
Thanks Mona, so the swing of the PI is determined by the loadline of the output tubes. Shunt/Shade NFB affect the load of the PI, so "mow much the ac loadline will be steeper than the dc load line", so how much current is needed to deliver the full swing to a(12ax7) = g1(EL8a), while mosfet driver is there to be able to supply more current and drive the output tubes into AB2.

In a way, yes.With more gain in the output stage the dynamic load on the pre get heavyer and it's gain drops.
Now I've understood, thanks!

Since the gain of the output stage isn't a constant the feedback load changes too, making the loadline of the ECC83 curved, no straigt line!
Are you referring to the passage from class A to the steeper loadline of class B?
 
Speaker don't have a constant impedance making a changing load to the final stage and consequentaly changing the gain witch reflects on the feedback.
Thanks again, so the only way to have shunt/Shade NFB and still have a straingt loadline is implementing it in an amp working in Class A with current feedback to achieve almost infinite damping (so that the variable impedance of the load won't affect the power tubes and so the PI/driver)?
 
Hi, may I ask how the PI's 220k 47k 16k has been chosen?

I understand that has been chosen a value to have:
- an high PI plate load to make the 12ax7 more linear (220k + 47k = 267 kOhm in this case);
- a PI plate load that guarantees a good working point at half the current set by the CCS;
- set a PI working point that guarantees good linearity at the chosen current;
- the 47k + 16k + 47k = 110 kOhm in parallel with the OT to be at least 10 times Raa in order not to waste too much power through it.

I would like to know if someone has tried different combinations and how it sounds in different configurations, EG:
- 220k 47k 16k (as it is);
- 180k 82k 27k (same dc loadline, steeper ac loadline, less wasted power);
- 150k 120k 39k (again same dc loadline, steeper ac loadline, even less wasted power);
- etc...
- 47k 220k 75k (extreme opposite case)

So, considering that 0,6 mA on the 267k loadline will make the triode working approximately at -1,5V, I would ask if other dc loadlines has been tested, EG:
- 100k 47k 16k (147k loadline, needs 1 mA to work at -1,5V considering 320V for the EL84s)

From 12ax7 datasheets ( http://tdsl.duncanamps.com/pdf/vm345.pdf ) it seems to have 10% less distortion at 1 mA than 0,6 mA.

Thanks in advance for anwering and sorry if what I ask is something obvious for most of you.
 
Before I begin my Baby Huey EL84 build I am looking at the PSU and OPTs. Bas Honeman has been a great help but I don't want to take over his build thread and this seems to be the best place for the question. As I am in the UK I need to make sure that I have 230v for the mains power. I need to order from the UK or EU before tariffs come in due to Brexit.

I may be able to order some Audio Note EL84PP transformers which are more expensive than these below from HiFI Collective UK Output Transformers | Hifi Collective

They would be on back order depending on Audio Note as to the lead times.

I have uploaded my BOM to Mouser UK but not included transformers.

There are some nice C Core OPTs from a site in Holland that Bas used but he has not answered my emails.

If I look at the UK

One option is from Primary Windings in the UK who provided the transformers for my EAR 834P Phone Clone.

510_T.jpg


Mains transformer Mullard 5-10 Stereo – 600V @200mA CT, 6.3V @4A and 6.3V/5V @ 2A – Primary Windings

'Designed to power a Mullard 5-10 Stereo amplifier using a single GZ34 or two EZ81 rectifiers.
Proudly made in the United Kingdom

Primaries
European primary voltages, 220V, 230V, 240V and 250V.

Secondaries
HT winding 300V – 0 – 300V at 200mA
Heater windings 6.3V at 4A centre tapped.

Rectifier winding 6.3V at 2A tapped at 5V at 2A.

Steel
High quality Grain Orientated M6 0.35mm.
Hand laminated for the best stacking factor, and accuracy.

Winding wire
Double coated Grade 2 enamelled copper wire.
Wound using CNC winders for highly accurate wire placement and turns count.

Electrostatic Screen
Solid copper sheet providing the best RF screening possible between primary and secondaries.

Connecting wire
High quality PVC 600V
Primary and HT : 22AWG UL1015
Secondary LT : 20 AWG UL1015

Weight
3Kg

Compliance
Built and tested to safety standard BS EN 61558-1 2005'



The Output Transfomers are here:
510_T_small.jpg


EL84 Push-Pull 10W Output transformer (Mullard 5-10) – Primary Windings
Product Number PWPP10W6K6-SU-08

A high quality 10W at 20Hz Push pull ultra linear output transformer. Designed for the Mullard 5-10 amplifier but may be used in any single ended design with a valve having an output impedance of 6600 Ohms anode to anode, such as an EL84.

Primary
Multi section winding, for reduced leakage inductance.

Inductance
Inductance 35H at 25Hz.
Leakage inductance 6mH

Bandwidth
Wide bandwidth, 10-50KHz. (-1dB)

Secondary
Output impedance 4,8 and 16 Ohms.

Steel
High quality Grain Orientated M6 0.35mm.
Hand laminated for the best stacking factor, and accuracy.

Winding wire
Double coated Grade 2 enamelled copper wire.
Wound using CNC winders for highly accurate wire placement and turns count.
Layer wound for excellent wind consistency and alignment resulting in reduced leakage inductance.

Connecting wire – stand up only
High quality PVC 600V
Primary : 24AWG UL1015
Secondary : 20 AWG UL1015

Mounting
Frame, Drop through or stand up versions, all with flying leads.

Weight
1.6Kg



Another alternative is from Oxford Electrical Products which are cheaper
s-l1600.jpg

OEP AUDIO / OUTPUT TRANSFORMER
OEP T29A05F - PUSH-PULL Audio /Output Transformer for EL84 Valve / Tube Amplifiers

Oxford Electrical Products (OEP) - design and manufacture in Oxford, England. By maintaining a close control over manufacture, OEP ensure that each transformer is manufactured to the very strict specification that their designer intended. The quality of manufacture directly governs the efficiency of the transformer for excellent sound quality and reliability.

APPLICATION - EL84 in PUSH-PULL mode. MULLARD 5-10

PRODUCT SUMMARY - Push-Pull output transformer (10W) for EL84 Valves / Tube application. Amplifier chassis mounting speaker / valve transformer

TRANSFORMER PRODUCT SPECIFICATION
CHASSIS Mounting
Maximum Power @ 20Hz: 10W

Impedance
Primary: 7k C.T. with 43% screen taps
Secondary: 4, 8 and 16 ohms
Frequency Response: (+/-3dB): 20Hz - 25kHz
Core Material
Grain orientated silicon steel
Mass: 1.4kg

Made in England, UK.

OEP AUDIO OUTPUT TRANSFORMERS
This transformer is part of the NEW specialist range of output transformers for SINGLE ENDED and PUSH-PULL GUITAR and HI-FI amplifiers. Please view our range of OEP audio output transformers for other applications - Other transformers may also be available, so please message us should you find the application you require is not shown.



Another Option that others have used are the Toroidy

Product code: 0B4B-668E0_20161214131751

TSTA 0250/001 - Mains transformer for tubes - Shop Toroidy.pl

aee275c86b88bd56d67f53ae1b6f6854.jpg


AUDIO grade and SUPREME transformers for amplifiers and other high-class tube audio devices.

Unlike commercial power line transformers, the AUDIO version is covered on the outside with a black Mylar tape which gives it an aesthetic appearance. The epoxy filled center with an opening for the M6 screw eliminates a need for an additional mounting bracket. The neoprene gum pad affixed to the bottom successfully eliminates transmission of vibrations.
The entire construction utilizes high-class, laboratory-tested parts. The core and the windings are impregnated, and the production process is being scrupulously controlled at every step.

Additional magnetic screening and the electric monitor of coated SYN-SHIELD copper tape between windings make the class AUDIO transformers one of the frontrunners in the area of safe use in the systems of electronic power supply.

SUPREME version is additionally shielded and potted in hand-made stainless steel box. Side-wiring mounting plate is also available.

We guarantee quiet performance and the highest electrical parameters.
Technical data
Nominal power 250VA
Input voltage 230V or 2x115V
Secondary SEC I 0 - 275V - 330V / 0,4A
Secondary SEC II 50V / 0,1A
Secondary SEC III 6,3V / 7A
Secondary SEC IV 6,3V / 7A
Secondary SEC V 6,3V / 2A
Insulation strengh 4kVA
Dimensions
(Audio version) Height 49mm, Outer diameter 119mm, Weight 2,5kg
Dimensions
(SUPREME version) Height 65mm, Outer diameter 150mm, Weight 3,4kg




The EL84 OPTs Product code:8788-947C9

TTG-EL84PP - Tube output transformer [8kOhm] 2xEL84 Push-pull or similar - Shop Toroidy.pl

Toroidal Push-Pull output transformer for tubes:

2xEL84 , 2x6V6 (2x6П6C) , 2x6BQ5 , 6P14P (2x6П14П)

introduces benefits of toroidal cores to tube output. Through the use of innovative, internationally patented winding technology, we have achieved exceptional electrical parameters, such as flat frequency bandwidth in acoustic range, very low THD distortion and high quaity factor.

Transformer is also avilable in stainless steel box potted EXPO version.


Technical data
Intended for Push-Pull
Core type Toroidal
Ultralinear tap 43%
Nominal Power 50W
Nominal anode current 200mA
Frequency bandwidth (-3dB) 8 Hz - 58 kHz
Secondary Impedance 4 and 8 Ω
Primary Impedance 8,0 kΩ
Turns Ratio (Np:Ns) 44,72:1(4Ω) , 31,62:1 (8Ω)
Primary Inductance Lp 404 H
Primary Leakage Inductance Lsp 4,33 mH
Total Primary DC Resistance 61,3 Ω
Effective Primary Capacitance 1,3 nF
Dimensions (standard) 115 mm (OD) x 65 mm (h), weight: 2,5 kg
Dimensions EXPO 120 mm (OD) x 80 mm (h), weight: 4,3 kg


Obviously we have EI transformers and Toroidial what would you recommend between these from above for my BH EL84?
 
Member
Joined 2004
Paid Member
The style of transformer is up to your taste and availability in the correct specs. I would buy them close to home, if possible, to reduce shipping cost on hefty magnetics and have any problems resolved locally, as well as keeping local suppliers in business.

If you used the power supply built into the PCB, the Baby Huey EL84 requires a power transformer with single 230Vac primary (not center-tapped) with 40-50vac for bias and sufficient heater current. (Study the heater arrangement in this thread - Marc recommends DC heater power). The power supply on Prasi’s board uses SS rectification to end up with 300-325 Vdc for B+. The power transformer from Primary Windings is center-tapped and intended for the 5-10 circuit which uses automatic bias. It is not ideal, but could be made to work (you probably will need to use valve rectifiers to get the B+ right and a separate bias supply transformer).

If budget is not too restrictive, my preference would not be the Primary Windings output transformer, nor OEP, the other UK transf you listed. Here is why.

1. I believe you should look for an output transformer as close as possible to the Dynaco Z565 specifications for your Baby Huey EL84 (BH) build. This design has stood the test of time and has never been advanced at reasonable cost. That means 8k primary to your preferred secondary, with 23-25% Ultra Linear taps, and a rated output of 17.5 watts. (Those were the specs Bas Horneman also asked for at Vlaartronics that you contacted).

2. Why 17.5 vs. 10 watts? In an optimized setting a PP pair of UL EL84/6P14Ps can produce around 17 watts. BH should well exceed 10 watts, although I have not seen actual measurments. So, to ensure good bass performance I will look for >15 watt output at your desired lowest frequency. I see the OEP transf is rated for 10 watts max AT 20Hz, which in practice is quite good. Some manufacturers obscure the lowest frequency achieved at their power rating.

3. I am currently building a BH EL84 using Toroidy 8k UL output transformers because I am curious to hear how they compare to my vintage Dynaco Z565 transformers, which are now in a modernized Dynaco SCA35 integrated amp. These Toroid transformers are not ideal for EL84 PP with their 43% UL taps, but they are rated at 40 watts and of very good quality at reasonable cost, so I could repurpose them for other power tubes if necessary.

Bottom line: You asked in a different thread about Toroidy chokes. If you buy the chokes, get the output transformers at Toroidy too. Unfortunately their power transformer TSTA 0250/001 is of too high voltage for the EL84 BH (but perfect for the EL34).

Hope this helps. Good luck.
 
Last edited:
Member
Joined 2016
Paid Member
Toroidy also makes custom voltage version at least of their power transformers at reasonable extra cost, just ask them for the voltages and configuration you need.

I have ordered from them several customized transformers already, and have always been satisfied.


By the way, the Toroidy EL84 PP output transformers are amazing. I'm currently listening to them ... although in a John Broskie (TubeCad) style EL84 PP connected in triode. Great little amp :D


Best regards, Claas
 
Member
Joined 2004
Paid Member
Toroidy also makes custom voltage version at least of their power transformers at reasonable extra cost, just ask them for the voltages and configuration you need.

I have ordered from them several customized transformers already, and have always been satisfied.

Best regards, Claas

Claas,

Thanks for your feedback on Toroidy. Good to know you are enjoying your EL84 output transformers.

Could you give some indication of the premium they charge for custom voltages in their power transformers? That would be great if they could make a power transformer just right for the Baby Huey at a good price.

Thanks,
Francois
 
Member
Joined 2016
Paid Member
Hi Francois,

sure - in my last order, the additional cost for changing secondaries from 2 x 10V to 2 x 8.5V was 10% of the transformer price, which came to less than 3 EUR / USD. I also remember sometimes voltage changes at no additional cost in previous orders.

The other transformer I ordered last time was a B+ transformer for a new PSU for my Aikido pre, where I asked for a 50VA Audio Grade with a single 250V secondary.
Standard voltages for the series I wanted to buy go to 230V primary / 230V secondary. This is no dedicated tube mains transformer, so I guess they had to build it to a different spec once the secondary voltage is more than 230V and higher than the primary voltage. Here the additional cost for higher voltage, for the Audio Grade transformer version, came to about 15 EUR / USD compared to the version at store-listed voltages.

For my different DIY PASS amps mains transformers (mostly 400VA / 500VA Audio Grade), I sometimes asked for longer leads, which led to only nominal surcharges, if at all.


I hope this helps,
Claas
 
Last edited:
Member
Joined 2004
Paid Member
Hi, may I ask how the PI's 220k 47k 16k has been chosen?

I understand that has been chosen a value to have:
- an high PI plate load to make the 12ax7 more linear (220k + 47k = 267 kOhm in this case);
- a PI plate load that guarantees a good working point at half the current set by the CCS;
- set a PI working point that guarantees good linearity at the chosen current;
- the 47k + 16k + 47k = 110 kOhm in parallel with the OT to be at least 10 times Raa in order not to waste too much power through it.

I would like to know if someone has tried different combinations and how it sounds in different configurations, EG:
- 220k 47k 16k (as it is);
- 180k 82k 27k (same dc loadline, steeper ac loadline, less wasted power);
- 150k 120k 39k (again same dc loadline, steeper ac loadline, even less wasted power);
- etc...
- 47k 220k 75k (extreme opposite case)

So, considering that 0,6 mA on the 267k loadline will make the triode working approximately at -1,5V, I would ask if other dc loadlines has been tested, EG:
- 100k 47k 16k (147k loadline, needs 1 mA to work at -1,5V considering 320V for the EL84s)

From 12ax7 datasheets ( http://tdsl.duncanamps.com/pdf/vm345.pdf ) it seems to have 10% less distortion at 1 mA than 0,6 mA.

Thanks in advance for anwering and sorry if what I ask is something obvious for most of you.

Hi dear Rorberto, please don’t take the “no response” as no interest. I had hoped to hear from more knowable members. Please keep this alive.