• 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

Member
Joined 2007
Paid Member
Jkeny,

*Concerning your question about using PS toroid trafos as OPT:

Take a look here: http://jjs.at/electronic/class_ab_subminiature.html
This guy designed a PP-subminiature tube amp using 5902 submini pentodes and toroid transformers.
He also gives some hints about the calculations for selecting the right toroid (pri/sec) voltage.
You should select a toroid core with a higher VA rating while using the EL84.
(The 5902 yield around 8W in Pushpull at 130V in my application with a Hammond 1615 OPT, he selected a 10VA model here)

And take another look here: http://geek.scorpiorising.ca/GeeK_ZonE/index.php?topic=3662.0
Here Tom Schlangen designs a small amp also with a PS toroid as OPT (take care: experimentation status)

My remarks to toroids:
- In the Baby Huey, a PS toroid is not appropriate as it is missing the UL-tap. The UL-operation seems to be vital
for the Baby Huey's magic.
- Toroids are sensitive to DC imbalance (generated by poorly matched EL84's). To circumvent this issue, keep the design
of Gingertube: one CCS per EL84 and each tightly trimmed at 38mA. If both tubes pull exactly the same idle current,
there's no DC-imbalance, hence no early core saturation of the toroid (= distortion).
- The people that use PS toroids as OPT's seem to use the Talema brand (here in Europe, Amveco in the US?). They are
potted in a blue plastic enclosure and are meant for PCB mounting. I got mine from Reichelt in Germany.

*Concerning your question about Blackgates:
I'm usually against "boutique" parts, partly because of the snakeoil, partly because of the price of the snakeoil.
But the Blackgates were the first caps that made a real audible difference to me when I soldered them in the cathodes of the EL84s.
They are worth their price. Take care: they need around 100hours to fully break in, for the first hours they sound awful.
(I'd didn't believe that either until i experienced it myself).
I was unable to measure any electrical difference compared to a std. electrolytic cap, but on the ear it's another story.

Kind regards,

Yves
 
Thanks Yves,
Good information!
- In the Baby Huey, a PS toroid is not appropriate as it is missing the UL-tap. The UL-operation seems to be vital
I've already built an ECL86 version - modified a Rogers Cadet III amp - and it used the existing OT which didn't have UL taps - magic was there in spades (and this was without global feedback)

- Toroids are sensitive to DC imbalance (generated by poorly matched EL84's). To circumvent this issue, keep the design of Gingertube: one CCS per EL84 and each tightly trimmed at 38mA. If both tubes pull exactly the same idle current,
there's no DC-imbalance, hence no early core saturation of the toroid (= distortion).
So do I need to get matched EL84s, as well as balancing them with CCS or will unmatched do & allow CCS to do it's thing?

I will give some Blackgates a trial run so!

Any ideas on the BDT phase splitter/vol control idea - is it a non runner because of low MU? - I'm keen to use it somewhere because it's so elegant & useful. It's the equivalent of an LDR attenuator for tubes!
 
Member
Joined 2007
Paid Member
Hi Jkeny,

So do I need to get matched EL84s, as well as balancing them with CCS or will unmatched do & allow CCS to do it's thing?

The CCS on the EL84 do the major part by keeping the idle current
constant and thus preventing core pre-saturation.
Matched EL84 pairs will help to lower the distortion in the dynamic condition (music playing).
Imagine unmatched pairs by having unequal amplification of the upper and lower part of the signal (it's a push-pull, each valve is responsible
for either upper or lower portion of the signal).
Matched pairs aren't absolutely necessary, if you buy some anyway then buy them matched.

I can't comment on the suitability of the BDT phase splitter here,
you'll have to go ahead, test and report here. The baby Huey design calls for a high mu valve run with low currents (hence the ECC83 choice). I guess that you could replace the ECC83 by two small signal pentodes (connected as pentodes) that have a suitable amplification and also run them with low current.
I admit that the beam deflection tube volume control idea is interesting, but John also said later on that this preamp had microphonics and some other limitations.

Kind regards,

Yves
 
gingertube said:
Diomedian,
You may well get away with a Hammond Classic Series 260G.
This is 275-0-275 High Voltage at 200 mA with Filament windings as well (6.3V @ 5A and 5V @ 3A).

You may need a choke after the rectifier or a couple of power resistors in the HV leads before the rectifier diodes to drop a little voltage.

Cheers,
Ian


Some newbie questions about power transformers-

I am trying to spec a baby huey power transformer and earlier in this thread there are comments about the 270HX 275-0-275 @ 200ma being a little too high of a voltage (post rectifier, I'm assuming). However, the schematic is calling out a 272HX 300-0-300 @ 200 ma. If the 270HX HT is too high, won't the 272HX HT be even higher?

If I want to build mono-blocks, do I simply spec the same voltage but 1/2 the current on all three windings? ie replace the 272HX with a pair of 272BX's and replace the 270HX with a pair of 270DX or 270EX?

Finally, are the any suitable off the shelf power transformers (from other mfrs) that are "Z mount" through the top deck style?? ie not not "X mount" stand-up mount?
 
boywonder,
For a stereo amp the 370HX will be fine - I would use a choke in the supply, that is, rectifier, small first capacitor (approx 2uf), then 2 to 8 H choke, followed by 220uF (what is called a CLC supply on this forum).
For improved result use separate chokes of 2H followed by 100uF for each channel after that first capacitor.
The output High Voltage (under load) can be adusted by adjusting that first capacitor between 1 and 5uF.

For Monoblocks use 370EX and use the higher value (8H) choke.

Use polypropylene capacitor for that first cap and parallel the electrolytic on the end with a polypropylene (or even polyester) 1uF cap.

BTW the 370 series trannies are the same as the 270 series but with universal primary windings.

The regulated supplies as in dupyv's post#369 can also be used. While I haven't tried it myself, I believe his statement that it improved stereo imaging and fine detail. Because the Baby Huey effectively uses only one High Voltage supply for the entire amp it is highly likely that an improved supply will give the benefits claimed.

Hope this is of some help to you.

Cheers,
Ian
 
I too am getting close to starting my "Baby Huey" build. I have picked up a Hammond 370HX power transformer. 550VCT 200mA, 6.3VCT 6A, 5VCT 3A, 50 V bias tap.

I would like to build this amp with valve rectification, so powering the CCS from the 5 volt secondary will not work. I see that in November, LowRedMoon (Adam) was asking questions about using valve rectification. GingerTube recommended that he use the 50volt (bias) tap of the transformer to provide the power for the constant current supply. I did not see if Adam followed this recommendation, so could not check on the implementation. As my electronics skills are rather rudimentary, I would be most grateful if someone could take the time to suggest the modifications necessary to the circuit to use the 50 volt bias tap as the power source for the CCS.

Thanks in advance,

Chris
 
Gingertube-thanks for the info above. I went for a pair of 270EX hammonds and a pair of 8H hammond chokes to build monoblocks. The 370EX was not readily available, and I don't really need universal AC input. I'm also assuming that I won't need the 50V winding on the 370EX for future mods...

I'm assuming that the above cap values for CLC remain the same for monoblocks, that is xfrmr, rectifer, 2mf, 8H, 220mf for each supply. Is this a good starting point?
 
To run the CCS from the 50 volt tap I was thinking of making a voltage divider, however, I don't have the experience/skill/knowledge to work out the amount of current draw that he supply will have to supply. This makes it a little difficult for me to work out the resistor values. As I wish to start the project, I have decided to go the "crack the nut with the sledgehammer" approach. I have a small 15 volt 5VA PCB mount transformer and a 7815 voltage regulator in the parts box, so I think I will make a simple regulated 15 volt supply. The original design has a voltage doubler off the 5 volt supply, giving about 14 volts. Any problems with my approach?

Thanks,

Chris
 
Chris,
Just use the voltage doubled 5V winding rather than fit a second transformer.

You will not need to worry about the 50V bias tap unless you want to build a version with fixed bias output tubes - in which case the -14 volts (actually probably a bit less than that) is too marginal for the bias supply and you need to generate a larger -ve rail using a single diode (cathode to the 50V tap, anode to -ve of a filter cap of say 100uF/100V, +ve of the cap to 0V) to get around -65 to -70 volts.

Cheers,
Ian
 
Thanks Ian,

My problem is that I would like to use the 5 volt winding to run the filaments of a valve rectifier, so I am looking for an alternate power source for the CCS. Looking at the schematics again I see that I have made a small mistake, the power supply is -14 volts, as you mentioned. I have a 7915 regulator also (-15volts). You mentioned last November to one builder that wanted to build valve rectification to use the 50 volt tap to supply the CCS since the 5 volt tap was being used to run the rectifier filament, hence my initial interest in this approach.

My limited electronic knowledge leads me to think that if I am running a valve rectifier, the B+ will be riding on the filament voltage. Is this correct? If so, I am assuming that running the CCS supply off this 5 volt filament tap is not going to work.

I have the parts in the parts box, the little transformer is only 40mm * 35mm * 35mm and I have plenty of room in my chassis to go with a simple regulated -15volt supply.

As I stated, my experience is limited, so not sure if I am pointing in the right direction ;)

BTW, when you mentioned in a previous post the stuff you were doing at work, combined with you living in Adelaide, made me think DSTO. I did some work with them in a previous life flying some of their projects in HS748s.

Cheers,

Chris
 
tube and circuit selection

regarding tube selection, the 6sn7 type tubes are very high quality, but for more gain, try the 6sl7 type of dual triode. for a novel approach, check out the 6bm8. ihave seen concern about power 'pentodes" (6k6) versus power "tetrodes". power pentodes have a larger plate and 3 physical grids, and the grids are alligned so they are in a straight line so that they shield one another...beam pentodes are sometimes mistaken for tetrodes because they have only 2 physical grids, but use beam forming plates to serve as a virtual suppressor grid, and so may be considered as a pentode class tube in most respects
 
regarding rectifiers; ive run across more than one 6bq5/el84 pp amp using a 6ca4 tube as arect...the isolated cathode and 6.3 volt/1amp fil of the 6ca4/ez81 rectifier make it a handy match, andby using a pair of 6ca4's with series fil's andplates in paralel can give considerable flexibility...put p1/v1 to p1/v2, p2v1 to p2/v2 , k/v1 to k/v2
 
Thanks for the interesting suggestions re the 6CA4s. I will admit that my interest in using a valve rectifier is one based largely on aesthetic rather than technical reasons. I have a couple of Soviet 5U4Gs and would like to use one for this project. I will, however, keep your suggestion in mind for future projects.

Cheers,

Chris
 
post377

regarding the dual-sheet-beam pentode-it dores still require a ballanced dual signal to function properly...i suppose that you might bias both grids at some point of about 2 to 4 volts, then apply signal only to one grid, but you could basically see the same thing with a long tailed pair...2 identical triodes, for example, sharing an unbypassed cathode resistor, both grids biased at the same voltage, a 1meg resistor between grids, a "large" cap from one grid to ground, the other grid receives the input signal. an appropriate matching plate resistor as plate load and then cap coupling to each grid of the next stage...i have seen this circuit used in the output stage having 6k6's, gibson used it with 6v6's, and i have seen it usud with 6L6gc's...remember, tho...the ONLY way you can use this arrangement is to bias the long-tailed-pair as class a!!!


you can also use an active splitter (class a) by using 2 triodes as a ballanced driver for a pp output by using appropriate plate load resistors and coupling the plates to the grids of a pp output stage...one triode gets a signal applied to the grid, the other uses a 50/50 between the output tube grids for its grid signal...the voltage divider provides considerable feedback to the second triode, as well as a near perfectly inverted signal which is very forgiving to any slight mismatch of parts...if you don't mind a bit more necessary drive, you can add a shared, un-bypassed cathode resistor to the phase inverter triodes...
 
post389

a 7812 or 7912 would give 12 volts...you could use diodes to drop the excess voltage...some drop .6volts, some as much as 1 volt... do NOT use zeners...they will catch fire unles used only as a voltage ref for an amplified (regulator) circuit. i need the tube types to calc. the current, but a bias winding generally puts out only an extremely small current...fils are usually the heaviest current draw...
 
I have about a dozen medical power entry modules (Schurter) that have the IEC power entry module combined with a power switch, fuse, and a medical RFI filter. Will this medical RFI filter on the AC input help or degrade the design? Should I just stick to a separate switch, cord entry module and fuse?

These medical power entry modules are tempting because I already have them and they are collecting dust....I can post a pic of the filter and the RFI specs when I get home, if required.

I bought a pair of 8H 100 ma chokes for the dual mono power supplies; however, now that I am hip-deep in Morgan Jones' book, he suggests sizing the current capacity of the PS choke for the post rectifier AC current as well as the DC current. Using this calc, it would appear that I need a roughly 150 ma rated choke for each supply....Is a 100 ma rated choke a problem here?

Finally, if I do xfrmr-rectifier-2mf-8H-220mf CLC, do I still need the 220K/2W resistor across the 220mf as shown on the schematic?
 
boywonder,
The IEC Power Entry Module with the RFI Filter will be ideal. I use these myself.

The latest "Baby Huey" monoblocks I finished last weekend use the Hammond 8H 100mA chokes. I'm still doing final tweaks on these monoblocks but the chokes appear to be adequate although 150mA rated chokes would be better.

The 220K/2W "bleed" resistor across the final 220uF capacitor is always a good idea. It makes sure that the capacitors are fully discharged when the amps are turned off so that (as long as you wait 5 minutes) you don't get nasty surprises when you decide to make a mod or some more measurements.

Cheers,
Ian
 
Baby Huey variations

Hi Ian,
I wondered, as a Baby Huey variation, that if a high current input/driver tube was used in the so-called E-linear configuration--ie connected to the UL OPT taps--could the Baby Huey cross coupled feedback thingy mechanism still be used? There will be two local feedback loops at work I guess, but I don't think that should be a problem.
As an extension of that "vapourware" (as AW calls it) what about using a pentode driver in E-linear with the screen (low current) connected to the output tube anode as per the BH? That at least would keep the screen voltage below the anode voltage. Then ya could even have two lots of BH feedback to play with.
Whatcha reckon 'cause I have no idea whether they would work.
best wishes
tim