SY -- want to dimension for me the size of the DPST switch that you are using to short the 100K input resistors? The ones I have from DK with the long bat-handles measure about 500 x 460 mils.
I think everything should fit without having to notch the board:
I think everything should fit without having to notch the board:
An externally hosted image should be here but it was not working when we last tested it.
I'll measure mine, but if yours are easily available, go with them. I used off-the-shelf from Radio Shack.
GND Lifting the Heaters with a 6.3Vac not CT-ed
I’m building this amplifier with different power transformers from those shown in the PS schematic (post #2) or used in the original design (Dyna SCA35). My problem really, is with regards to the filament wiring. I have a straight 6.3Vac winding (i.e. without centre-tap) to which all the valve filaments will be connected in parallel. Now, the ECC81 driver/phase-splitter valves must be floated above GND.
I am thinking of series connecting 2 x 100R || 10nF across the 6.3V points and from the junction make a connection direct to the GND lift network on the B+2 PS output. I believe this should be fine for the ECC81s, but how do I connect the 4 x EL84s to the same 6.3V transformer (ie in parallel with the ecc81s) without lifting them also from GND?
I hope I’ve made myself clear and would appreciate help from other builders and/or competent diy-ers.
Thanks,
Joe A
I’m building this amplifier with different power transformers from those shown in the PS schematic (post #2) or used in the original design (Dyna SCA35). My problem really, is with regards to the filament wiring. I have a straight 6.3Vac winding (i.e. without centre-tap) to which all the valve filaments will be connected in parallel. Now, the ECC81 driver/phase-splitter valves must be floated above GND.
I am thinking of series connecting 2 x 100R || 10nF across the 6.3V points and from the junction make a connection direct to the GND lift network on the B+2 PS output. I believe this should be fine for the ECC81s, but how do I connect the 4 x EL84s to the same 6.3V transformer (ie in parallel with the ecc81s) without lifting them also from GND?
I hope I’ve made myself clear and would appreciate help from other builders and/or competent diy-ers.
Thanks,
Joe A
That means of creating an artificial CT should work fine.
As long as the EL84 heaters are within their heater-to-cathode limits (which they are in this design), there's no problem lifting them from ground.
As long as the EL84 heaters are within their heater-to-cathode limits (which they are in this design), there's no problem lifting them from ground.
I'm in the first-testing phase of my RLD 84 build (as per SY's article) but I'm getting ridiculously low voltages (4 volts) on the input valve plates and phase-splitter plates and cathode. This voltage is taken without the output tubes. The voltage on the respective plate resistances is OK - around 380v. The voltage from the screen regulators is also OK - around 285v - but these are obviously out of the ecc81 circuit.
I also measured the heater-float voltage and again the voltage is similar (in the 10v range). Again the voltage prior to the heater float RC network is 380V. I made resistance checks for the OV line of the amplifier circuit (but not, I am now realising, for the power supplie(s) ground).
Does anyone have an explanation for such low voltages at the valve plate pins?
I also measured the heater-float voltage and again the voltage is similar (in the 10v range). Again the voltage prior to the heater float RC network is 380V. I made resistance checks for the OV line of the amplifier circuit (but not, I am now realising, for the power supplie(s) ground).
Does anyone have an explanation for such low voltages at the valve plate pins?
If I understand correctly, this is just for the input tubes? Do you get the same result with the input tubes removed? If you used the protection diode on the phase splitter grid-to-cathode, make sure it's not backwards or shorted.
Not so much as you'd think. If the diode is shorted or backward, the max current that can flow is the B+ divided by the sum of plate and cathode resistors. That's still only a few milliamps so a fuse won't save the day.
This is why I love, love, love my variac.
This is why I love, love, love my variac.
Hello SY, I have no diode on the phase splitter grid-to-cathode. I just built the amplifier as shown in your article on Bas' website.
Yes, this is just for the input tubes (i.e. without the el84s). With (all) the valves removed I still get the same low voltages.
By the way, dsavitsk, my 1 amp fuse (on 230Vac mains supply) survived several turn-ons/offs, if that's what you mean.
Thanks for your help.
You might want to check the latest version, which is on my website. No significant differences for the original version, but there's a few minor tweaks. In any case, pull out the driver tubes and recheck voltages.
What could be wrong with my PS?
This is really a continuation of my previous emails. At the moment I stuck trying to get the correct heater-floating voltage, and all I’m getting at the 200K – 51K junction is 5Vdc! I’ve already spent many hours replacing/breadboarding/testing all components from the end of the choke onwards and each time I get these low voltages. It’s been all very frustrating, so I’m turning again to the forum for help!
I’m using one transformer with 0-270-310 volts AC for all B+ points. The 0-270V powers the output valves (EL84s) giving me around 320V on the plates (with OPTs in place) which is OK; the 0-310V taps power the ECC81 drivers/phase splitters and the heater-float dividing network. B2+ is around 400V. The curious thing about it is that I get very low voltages at the end of every resistor that is fed by B2+. This is the case at the heater-float junction (as already said), at the plate of the ecc81 driver (5V), at the plate of the phase splitter (12V) and at the cathode of the phase splitter (5V). I have not used the protection diode on the phase splitter grid-to-cathode.
Can someone please help explain what could be wrong in such a simple circuit? Thanks in advance!
This is really a continuation of my previous emails. At the moment I stuck trying to get the correct heater-floating voltage, and all I’m getting at the 200K – 51K junction is 5Vdc! I’ve already spent many hours replacing/breadboarding/testing all components from the end of the choke onwards and each time I get these low voltages. It’s been all very frustrating, so I’m turning again to the forum for help!
I’m using one transformer with 0-270-310 volts AC for all B+ points. The 0-270V powers the output valves (EL84s) giving me around 320V on the plates (with OPTs in place) which is OK; the 0-310V taps power the ECC81 drivers/phase splitters and the heater-float dividing network. B2+ is around 400V. The curious thing about it is that I get very low voltages at the end of every resistor that is fed by B2+. This is the case at the heater-float junction (as already said), at the plate of the ecc81 driver (5V), at the plate of the phase splitter (12V) and at the cathode of the phase splitter (5V). I have not used the protection diode on the phase splitter grid-to-cathode.
Can someone please help explain what could be wrong in such a simple circuit? Thanks in advance!
I'll repeat my earlier question- if you pull out the driver tubes and measure the voltage at the plate pins (1 and 6), is it still 5V?
Yes, SY, with the driver tubes removed the voltage at the plate pins are as reported earlier: 12V at the plate of the splitter and 5V at the input/driver plate pin. Thanks a lot!
Joe A
Joe A
Remember, a transformer transforms current -- so the idea of an artificial center tap may not work -- I think that this is why the little Dyna trafo didn't work well with the Impasse, but the unit from Allied works like a champ.
SY -- what did you "tweak" in the newest version? Is it sufficiently tweakified to qualify as a POOGE?
SY -- what did you "tweak" in the newest version? Is it sufficiently tweakified to qualify as a POOGE?
No, nothing like that. Since the original article, I dropped the diode strings to six in series and added small current balancing resistors. Of course, if someone wants to use co-poly(ethylene norbornene) capacitors or sputtered rhodium resistors, have at it.
As soon as I get a little more time, I have some ideas for a redesign using some local feedback around the output stage and CCS loading in the first stage, but that would be a bit beyond "tweaks."
As soon as I get a little more time, I have some ideas for a redesign using some local feedback around the output stage and CCS loading in the first stage, but that would be a bit beyond "tweaks."
The main disadvantage is that the 6V6 takes a higher bias, so the LED strings will be proportionately longer. That means higher parts count and a somewhat higher dynamic resistance. Still probably less than the customary 10R cathode resistors, though.
