• 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

I have just started the build process for my "Baby Huey". This is my fifth valve amp build, but the first this complicated, with constant current circuits. I think I am reasonably comfortable with wiring things up now, but would like to follow a step-by-step process of building and checking as I go. I have built the bias blocks and constant current sinks on perf boards with screw terminal connectors to make them modular, but was wondering if anybody could offer some advice on testing them. I would prefer to make sure they work before just connecting everything together and doing a "smoke test". As I am building this with a valve rectifier, I have built a small -15 volt regulated supply for the current sink (tested and working).

Any input/advice greatly appreciated!

Thanks,

Chris

CCS and Bias Blocks

An externally hosted image should be here but it was not working when we last tested it.


-15 volt supply

An externally hosted image should be here but it was not working when we last tested it.
 
Chris,
These look good.
To test the -15V supply hook up a 1K resistor across the output and check that you have 15 volts across it. That won't check it to full capability but will give you a confidence test.

To test the diff amp current sources hook them up to the -15V supply you made and connect a dummy load of say 4K7 from the upper transistor cathode to 0V. Then check that you are developing approx 4.7V across the 4K7 (that is approx 1mA through the 4K7 dummy load).

To check the output tube cathode CCS (one at a time) hook it up across the -15V supply with the PS -15V to what would be the amp 0V connection and the PS 0V connection to what would be the amp cathode connection (to get polarity right) Then check for 0.6 to 0.7 Volts across the 16 Ohm resistor.

If you can be bothered I would be tempted to rebuid your -15V supply into a +/- 15V supply so that you can do the Yves (dupyv) mod shown in the top left corner of the schematics at post #369. This is just returning the 5K6 resistor (R1) in the EL84 CCS's to +12V (+15V in your case but its not critical) instead of wiring it to the tube cathode and the top transistor collector. Its not absolutely necessary but it will improve the stability of the CCS particularly under heavy drive of the amp - at the expense of requiring a +ve low voltage rail. Yves seemed to think it was an improvement and theory says he's most likely right.

I would not suggest you tackle his other mods but that one is relatively easy if you can be bothered redoing your -15V supply to +/- 15V.

Cheers,
Ian
 
Thanks for taking the time to respond Ian. I don't think I will bother with the +15 volt mod as I am building this amp while I should be doing uni assignments ;) can't waste too much time... I will keep this in mind though for the next amp, as I am thinking of a PP 6L6 next, or I might mod this one in the future.

Thanks again, will keep posting as I progress.

Chris
 
Thanks Ian,

Your assistance helped me to identify a minor wiring error in one of the bias blocks. Managed to fix it without damage to any of the components.

All Bias Blocks are reading 6.4 volts (I am using 18R resistor as I expect a B+ a little on the higher side).

CCS for the input triodes are reading 6.8 volts across a 4K7 resistor. This works out to 1.45mA, a little on the high side. My question, is this OK? Looking at the ECC83 curves it looks like 170 volts and 1mA are -1.5v, while 1.45mA will place it at about -1.1v.

Should I leave as is or any suggestions to adjust?

Thanks for the ongoing help! :)

Chris
 
Hmmm,

Feel like a bit of a dill. Was wondering why the different result than expected from the CCS. Noticed that the LEDs were not lighting. I had them wired backwards :(

Re-wired. LEDs light, voltage reading is 6.03v which works out to 1.28mA. Not quite 1.0mA, but closer than with incorrect setup. Hopefully this is close enough now.

Cheers,

Chris
 
It's the voltage drop across the LED's diode junction minus the base-emitter drop of the BC547/549 that sets the voltage, and hence the current, across the 1Kohm resistor. There will be some variation in the LED forward voltage drop between LEDs as well as variation due to current through the LED. There will be some slight device-to-device variation in the base-emitter voltage in the '547/549s as well. You could breadboard and match LED/transistor combos or change the 1K resistor to a 1.1 or 1.2K and put a trim pot (say 20K) across it, if you want to be anal about achieving exactly 1 mA. All that being said, I don't expect that the 1.28 mA will be a problem. The CCS as designed is rather temperature sensitive anyway so you won't know what the final actual current is until it's in the amp and warmed up.
 
I have started the construction. Most of the chassis work is done, just have to drill the back panel for the speaker binding posts and drill hole in front panel for the power switch. Hope to start wiring tomorrow. Want to do the filament wiring first, but have run out of single core of sufficient gauge.

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 
Time to bore you with some more progress shots.

In previous projects I wired the heaters with 1mm single core wire. With this project I purchased valve sockets that have small posts rather than eyelets. They look good, but I soon discovered that wiring he heaters with 1mm wire was going to be impossible. Only other single strand I have is 23AWG, a shade under .6mm. Mr. Jones states in "Building Valve Amplifiers" that .6mm is OK for 9 pin valve heater wiring. He did not elaborate on how many in the string can be wired with .6mm. I ended up wiring each valve heater with its own run of wire from the tag strip near the power transformer.

An advantage of having a -15 volt supply is that I could easily use a switch that has an led back light.

Hope to have it finished next week!

Chris

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


And, of course, the compulsory glowing valves shot...

An externally hosted image should be here but it was not working when we last tested it.
 
Bah !

Having each heater powered separatly is not a bad thing and your wire is probably large enough for that.

Easy to check remaining voltages !

But, why using small caps (?) at each socket ? Are there locally grounded ?

I can see the "usual" two resistors on the solder lug where the heater come from, no need for more cosmetics !

Yves.
 
Yves,

I recall reading in one of Morgan Jones books where he mentions using a small cap from each heater pin to chassis. Did this in my last build and it did not appear to do any harm. I am new to this, so if I am doing the wrong thing I can always remove them. My first point to point effort hummed and buzzed so I went in to the books and tried to absorb as much as possible to make a quiet amp. The last two builds have been hum free, but I am a bit paranoid after that first build...

boywonder,

The switch is something I picked up one day while browsing at an electronics shop in Perth WA (Altronics). It is alternate action, DPDT, 12volt for the light. Think it was $8 or $9 here in Australia. Noticed that Rob11966 used a similar thing on his beautifully built Dynaco clone, that is what made me pick up the switch when I saw it, hoping to incorporate it in a future project.

Thanks for the interest,

Chris
 
Here is a link to the Altronics web site http://www.altronics.com.au/index.asp?area=item&id=S0937

The above mentions that the switch is rated at 500mA 250VAC, however the actual switch has 3A 220VAC printed on it.

The switch is ONPOW LAS2, google showed this http://www.made-in-china.com/showro...o/China-ONPOW-Pushbutton-Switch-LAS2-GQ-.html
However I suppose that you don't want to order 1000 units (min order) ;)

Surely you must be able to find something locally, you guys have everything over there!

Cheers,

Chris
 
LTP-CCS --Bias current

Hi, Gingertube

I am in the middle of building this amp and need to clarify the resisitor value controlling CCS bias current.

In the circuit, the RED LED gives 1.7V, and BASE-EMITTER junction takes 0.7V, this leaves the voltage across the 1K ohm resistor just about 1V, so the CCS has constant 1 ma cuurent flow..this means each side of the circuit has 0.5ma.

Could this 0.5ma be a bit low for 12ax7 ??
 

Attachments

  • ltpccs.gif
    ltpccs.gif
    19.4 KB · Views: 1,357
See the test result from Chrish above. He got 1.28mA

There is a trade off between getting the 12AX7 current higher and the anode voltage dropping due to increased voltage drop across the anode resistors.

About 1.2mA (that is 0.6mA each side of the diff amp) is a good value but 1.0 or 1.1mA would be OK.

If the anode voltages on the diffamp are less than say 140 Volts then decrease the CCS current a little. If the anode voltages are above say 190 Volts then you can increase the CCS current a little.

The way to think about this is ( for example)
The Output EL84s are going to need +/- 12V signal swing, so that is what has to appear at the differential amplifier (diff amp) anodes.

The feedback voltage appears 180 degrees out of phase at the other end of the 220K anode load resistor. Say we have +/- 24 Volts of feedack signal at that point (depending on the size of that cross coupled feedback set resistor). We would then need to develop a total of +/- 36V across the 220K load resistor which means we have to swing 36/220K = +/- 0.164 mA current through each triode of the diff amp. In that case 0.5mA would be OK - the current would swing down to 0.336mA and up to 0.664 mA. Output may be a little cleaner (less distortion) with 0.6mA idle current as it doesn't swing as close to 0 current where the worst non-linearities are BUT the distortion is mostly second harmonic and the diff amp action will take most of that out, I don't think it will make a lot of difference.

So while you will nearly always see 12AX7 operated at higher idle currents than this (typically more like 1mA) in this circuit 0.5mA to 0.6mA is fine and that helps keep the anode voltage at reasonable levels.

From the above you can also see that it is possible, if you use really high values of resistor for that cross coupling feedback set resistor, that eventually the 12AX7 would try to swing down to 0 current. I've found that about 47K or 56K are the practical limits on its value but I never liked the sound with any of those really high values. Around 15K to 20K sounds good - up to 33K can sound OK depending upon how "tight" and "neutral" you want the amp to sound. At 47K the amp sounds transistor-ish (those 3 legged fuse things) and is quite dull, sterile and boring.

Hope this "explaination" helps.

Cheers,
Ian
 
Thanks for the explanation Ian, and for the question jueic. When I plotting the load lines for the input triode, I forgot that each would see half of the load (doh!). No wonder I was scratching my head about 1.28mA being too much!

Hope to have my amp wired up tonight! Will post results if successful.

Cheers,

Chris
 
Thanks, Ian & Chrish

Now the circuit makes a much better sense to me after reading your explainations... I finally settle with 0.65ma and 12ax7's plate sitting at 140V. Music sounds quite good (and right) at this point.

The circuit is brilliant.. and very sensitive to component change, I can hear the difference of music presentation for each component change.. I am totally amused !!!

JueiC