• 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

Jump to Latest
#1,361
Hi Gingertube,

Thanks for the explanation. Just a couple more questions (i'm still learning).
1) Why use a 56R in the 6V6 shared cathodes? Shouldn't the grid bias provide the means to adjust output tube current?
2) What is the role of the 47k in the ccs for the FET?

Cheers ANdrew
 
#1,362
The 56R in the 6V6 cathodes is not there to provide any part of the bias although it will contribute to the bias for the 2 tubes. It is there to provide some common mode feedback to the output tubes which helps reduce 3rd harmonic and other odd order harmonic distortion and also reduce intermodulation distortion. This was discussed earlier, see post #451 and onward. In particular note Sheldon's spectrum analyser plots.

The 47K is an essential part of the "ring of two" current source (CCS). It provides the base current for the current "pass" transistor and the collector current path for the "control" transistor.

Cheers,
Ian
 
Last edited:
#1,363
gingertube said:
... It is there to provide some common mode feedback to the output tubes which helps reduce 3rd harmonic and other odd order harmonic distortion and also reduce intermodulation distortion.
Click to expand...

This is the method I used to reduce intermodulation with power supply ripple:
Replace the feedback resistor with two equal resistors in series so that the total resistance stays the same.
Now we have access to the 'virtual ground'. Measure it, it has all the power supply ripple and common mode signal from output stage.
Tieing it to ground with an elco (and perhaps smaller cap in parallel) will reduce the common mode signal reaching the differential stage. Enjoy the sound 😉

PS: hope this hasn't been mentioned before in this long thread.
 
#1,364
Baby Huey PSUDII Power Supply model with 5AR4 (long)

[image][/image]In the image below I've modelled the power supply using the 5AR4 rectifier. When I first modelled it, I got an error saying the current forward maximum of the rectifier (IFRM) would get exceeded at a certain time. Then when I changed the model to a 5AR4-Graetz the issue was no longer present. I already have in my spare part box some 600V (6A) Cree Shottky diodes.

http://www.diyaudio.com/forums/gallery/showphoto.php?photo=8568


Question(s) :
1) Is the 6A 600V diode safe enough to use in regards to voltage for rectifier use? ( I already have them)

2)If so ..can I place a single Cree diode in series with each plate (pins 4 & 6) of the 5AR4 rectifier and still use the centre tap of the mains transformer?

3) Or should I not use the centre tap and tie the diodes together something like this:

http://www.lundahl.se/sidor/hybrid_power.html

I know that I could only use SS diodes and not be bothered at all with the rectifier but I would like to try this and it is a good learning tool for me. At some point I will revert back to only SS diodes for rectification but I would like to see for myself if there is a difference in sound and or operation.

Thank you in advance & Happy Victoria Day Weekend
 
Last edited:
#1,366
Found partial answer

Kevin Kr wrote :The 5AR4 is also noted for hot start arcing issues, particularly current production and placing a UF4007 in series with each plate completely prevents this issue and does not otherwise alter rectifier performance wrt voltage regulation, warm up delay or EMI.


(http://www.diyaudio.com/forums/tubes-valves/208717-diodes-graetz-hybrid.html#post2944465
)

So if I am correct in my interpretation of the above, I can place a UF4007 in series with each plate and still use the centre tap of the mains secondary as normal for a 0V reference?
 
#1,371
If it is a non center tapped HV winding then yes.
If it is a centre tapped winding then no.
You need a tube diode element in each [winding to rectifier to storage capacitor] circuit so that you can not get SS diode switching noise from reverse conduction as the SS diode switches off.
As the voltage on a SS diode reverses, you get a "splat" of current in the reverse direction which is basically sweeping out the charge of the SS diode junction capacitance. Once that charge is swept out then diode is then properly OFF and no current flows. It is that "splat" of reverse current which IS the SS diode switching noise. It couples horribly back through the transformer to any other windings. The tube rectifier does not allow that reverse current "splat" to happen.

ASIDE: The difference between standard SS diodes and the Ultrafast / Soft Recovery types is that that junction capacitance charge is 5 to 10 times smaller (so the reverse current "splat" is 5 to 10 times smaller too). Schottky diodes junction charge is 5 times smaller again.

Cheers,
Ian
 
#1,372
Ouput and power transformer advice

I think this is going to be a great project to get a friend into tubes 😀.

The project Wiki is excellent, but I have some questions. I've read 30 pages or so of this thread, and now my brain is about to pack up and leave 😉. I tried the search function but got scattered results. Accordingly, I apologise if my questions have already been answered.

Getting ahold of transformers here can be difficult. One supplier does bring in a limited range of Hammond stuff though. Is the Hammond 1608 good enough for the OPTs? I can also get the 1620 and 1650H.

As for the power transformer, I'll probably have to get toroidals wound locally. Should I consider getting 3 separate transformers, or one with 3 voltages? What are the individual VA requirements for 5V, 6.3V and 330(?)V

Thank you everyone!
 
#1,373
Hammond 1608 are the Output trannies I'm using on the Baby huey on my HiFi Syatem.

Here is something I posted earlier:

"In my on-going efforts to "convert" the youth of the world to Hollow State my apprentice at work has just finished building a fixed bias version of the Baby Huey for his home HiFi.

He wanted to use a single power supply for both channels and use readily available Hammond Trannies.

Here is what we came up with for his amp:

Power Supply (for 2 channels).
Power Transformer - Hammond 370HX (275-0-275 V @ 200mA).
Solid State Rectifier Diodes (Ultrafast Soft Recovery)
"C-L-C" arrangement power supply.
10uF/900V Polypropylene first capacitor.
Hammond 158S (1.5H 250mA 60 Ohms DCR) choke
470uF/450V final electrolytic with 5uF/900V polypropylene in parallel for the 2nd capacitor.
This gives a rock solid 330V rail - just right.

Hammond 1608 Output Trannies.

Notes:
The 900V polypropylene for the first capacitor is an overkill. You don't need a fantastic quality cap in this position.
The Ripple Current for this capacitor (according to PSUD2) is 320mA RMS
A Panasonic ED 10uF/450V has a 350mA RMS ripple current rating and could be used instead.
BUT
I would instead RECOMMEND using 2 off Panasonic ED 22uF/200V in series, with 220K voltage share resistors (1 or 2 watt) across each capacitor. These caps have 600mA ripple current rating which means a far longer lifetime.

For the cap in parallel with the 470uF/450V second capacitor, any 630V rated film cap of 2.2uF or 4.7uF would do. Polypropylene for preference but even polyester will be better than no parallel cap.

This supply would suit either the original (CCS biased) or the fixed bias versions of the Baby Huey."

Hope this helps

Cheers,
Ian
 
#1,374
My first scatch build ... so far

IMG_1076.JPG


Using plywood as a bread-board of sorts for now. Still learning as to the best placement as to where things go. Will move to a metal enlcosure. Plan is to have a separate PS and feed separate mono-blocks via umbilical as my shelf space is very limited as to where I can place things.

Learning lots as I go as this is my first scratch build & it is really slow going but lots of fun.
 
#1,378
Welcome said:
Quick question (and deal-breaker in my case) - can this easily be triode strapped or otherwise? My transformers cannot be used for ultra-linear, and I'm not experienced enough in this field (yet) to make modifications of my own.
Click to expand...

The design already has local NFB included, and Ultralinear looks like overkill here. I'd take Yvesm's suggestion to do without and run straight pentode mode. If you're feeling adventurous, and have an OPT with a multi-tapped secondary, you could consider cathode feedback instead.

Neither is strictly necessary, as you can always adjust the gNFB to compensate for the missing Ultralinear NFB, if that's even necessary.
 
You must log in or register to reply here.
Top Bottom