diytube.com

cedricb wrote:The 240V tap produced obviously a lower centre tap secondary (300V), but I've got 6V instead of 6.3V.

Any comments please?

Did you install R36 & R37? Their only purpose is to lower the heater voltage a little bit. If you put them in and the heater voltage is lower than you'd like, then just take them out and replace them with a jumper wire.

Ty_Bower wrote:Did you install R36 & R37? Their only purpose is to lower the heater voltage a little bit. If you put them in and the heater voltage is lower than you'd like, then just take them out and replace them with a jumper wire.

I've installed both jumpers from the start.

...and about these to reduce the centre tap secondary?

https://taweber.powweb.com/store/vdump.htm

or this at bottom of the page http://www.geofex.com/Article_Folders/mosfet_folly/mosfetfolly.htm

6VAC isn't really too low for the 6BQ5 / EL84s, IIRC the Russian tubes are rated 5.7 to 6.9V filament operation.

The Trafomatic "ST35" power trannie I have has two 115VAC primaries, and the standard USA voltage is 120VAC, and at my previous (NH) location varied between 123 and 128VAC

At my present (NC) location, it varies from 120 to 122 VAC.

/ed B

EWBrown wrote:6VAC isn't really too low for the 6BQ5 / EL84s, IIRC the Russian tubes are rated 5.7 to 6.9V filament operation.

...and about V1 (5751) and V2 (12BH7A) for the 6VAC?

...so is the 240V tap has resolved my voltage issues?

Should I bias at 350mV or lower with that configuration?

Back when they came out with the 5751 and 12BH7A, the voltage going to people's houses varied from 110 to 125. If a radio/amp/TV was setup for 117 to provide 6.3, at 110 there would be 5.9 VAC on those filaments. 6 volts will be OK, except for a very weak tube.

I've done some additional measures with the 240V primary tap.

V1: between 4 and 9 --> 5.86VAC
V2: between 4 and 9 --> 5.86VAC
V3: between 4 and 5 --> 5.86VAC
V4: between 4 and 5 --> 5.80VAC
V5: between 4 and 5 --> 5.93VAC
V6: between 4 and 5 --> 5.97VAC
On J1 between 7 and 8, I'm getting around 6.08VAC. So I don't know why there is a drop on the filament A and B...

Anything I can do to check that V1 and V2 are performing all right with the lower voltage filament?

On the schematic where it's saying 320V: I'm getting around 318V. Is it this value which was too high with the 230V primary tap? ...I didn't check again otherwise I need to re-bias...

Can someone explain the biasing in relation with the different voltages from the power tubes please? pin 3, pin 7 and pin 9

The four EL84s and the VA & PI tubes consume a total of around 3.7 amps, and the slight amount of resistance in the PCB's copper etches will "lose" a few tenths of a volt by the time it reaches the filament connections.


The biasing voltage on the cathode (pin 3) sets the total current passing through the tube, which is a combination of the plate and screen grid currents, which is easily measured at the cathode, since that carries the total of the two curents. The grid voltage is always negative in relationship to the cathode voltage, whether in cathode, or in fixed bias situations.

Typically, with a B+ around 365VDC, a cathode-to-G1 voltage differential of approximately 14VDC will operate the tube at or near 35 mA current, this voltage is not a set-in-stone "golden" value, but an average figure. I've seen EL84s and 6P14Ps which can vary from as low as 10VDC to as high as 16VDC to achieve 35 mA with 365 VDC B+ (in a typical ST35 design).

I've had matched quads of J/J EL84s which were dead-on at 35 mA, with my initial pre-setting of the cathode resistor networks to 400 ohms from pin 3 to ground, and then the "generic" Russian 6P14Ps can vary to the extreme high and low voltages (16V and 10V, respectively).

(I suspect that is how Sovtek and other tube vendors "grade" their tubes for guitar amp usage, the hotter tubes give better "grunge" , the lower gain ones are cleaner)

That would translate to cathode resistor values as low as 285 ohms, and as high as 457 ohms. THe "10V" tubes have higher gain than do the "16V" tubes.

I save these, "extreme voltages" tubes for SE projects, and as voltage regulators, and keep the 13V to 15V" tubes for ST35 / SCA35 usage.

The plate and scren grid (G2) voltages will be determined by the DC resistance of teh output transformer primary windings at teh plate and SG taps, and the tube's actual operating current.

35 mA isn't an absolute, ideal value, it is determined by the actual plate voltage and B+ , for lower B+ voltages (300 to 325VDC) the current can be up to and over 40 mA, and for higher voltages (380 to 400VDC) it should be set lower, around 32.33 mA, in order to stay within the safe maximum PD ratings (typically 13W total). With fixed bias operation, the total current per tube can be set lower, 25 to 30 mA.

For Class A SE operation, a constant currernt sink (CCS) can be used, to set the (cathode) current at an fixed value, and then any random EL84 / 6P14P / 6BQ5 can be used and it sill operate at teh pre-determined current. CCS can be designed using LM317, TL783, 7805, etc r three-terminal egulator ICs.

/ed B

Thanks EWBrown!!! I'll leave it for now and enjoy the music...

I've biased @ 345mV; these give me 13,6V for V3/V5/V6 and 12.6V for V4.

I'd just let it run for a while (a couple weeks' normal usage) and then re-check the bias and adjust if needed. With the 240V primary taps, you should be able to set to 35 mA for each EL84 / 6BQ5 and stay within the recommended PD ratings. THis also will allw the caps and trannies to settle / age in and the sound quality should gradually, and subtly improve with additional usage. I've found that most "newborn" amops can sound rough or shirll, but they (just like wine or whisky) will improve with age.

Enjoy!

/ed B

cedricb wrote:Thanks EWBrown!!! I'll leave it for now and enjoy the music...

I've biased @ 345mV; these give me 13,6V for V3/V5/V6 and 12.6V for V4.

Typically, with a B+ around 365VDC, a cathode-to-G1 voltage differential of approximately 14VDC will operate the tube at or near 35 mA current, this voltage is not a set-in-stone "golden" value, but an average figure. I've seen EL84s and 6P14Ps which can vary from as low as 10VDC to as high as 16VDC to achieve 35 mA with 365 VDC B+ (in a typical ST35 design).

I'm looking at a data sheet for 6BQ5 that show @300v w/no signal 35ma on the plate. With a 10v RMS signal the plate current jumps to @45ma. At 385v that's going to be @ 17W PD. No?

It doesn't work that way 20-20. Yes the plate current increases as signal is applied to the grid, but in doing so, the AC component of the plate voltage causes it to drop, such that at full power output, peak plate current might be well over 100 ma, but since the plate voltage has effectively dropped to something less than 100 volts (due to the signal developed there), plate dissipation is not exceeded. The simple way to look at this is that plate dissipation at any given time is equal to DC power in, minus the AC power output developed. Hope that helps!

Dave

dcgillespie wrote:It doesn't work that way 20-20. Yes the plate current increases as signal is applied to the grid, but in doing so, the AC component of the plate voltage causes it to drop, such that at full power output, peak plate current might be well over 100 ma, but since the plate voltage has effectively dropped to something less than 100 volts (due to the signal developed there), plate dissipation is not exceeded. The simple way to look at this is that plate dissipation at any given time is equal to DC power in, minus the AC power output developed. Hope that helps!

Dave

Much! The fog is lifting a little more every day. So the drive signal is always going to be seen inversely at the plate and doesn't cause PD to change from what is set by the biasing. (I really need to get my scope out from under the table). Once biasing is set under peak PD a drive signal cannot cause a Power Dissipation problem other than grid current on the positive swing if the signal is too high.( And cuttoff clipping ..etc)

Well, that's almost correct -- there's one other factor that can have a huge effect on Pd under full power conditions: Loading conditions can greatly affect Pd as well. As a gross example to make the point, let's say the output tubes are properly biased within ratings under quiescent conditions. With a proper load on the amplifier, more current will be drawn as signal is applied, but power will be transferred into the load, so relative Pd remains within ratings. But now, lets change the load. Let's short the output terminals. Now we're pulling crazy current, with no load to absorb it -- except the plate! That sucker will be glowing cherry red, even though the bias is properly set. So while the correct amount of bias keeps the Pd in check under quiescent conditions, the load placed on the amplifier has by far the largest effect on whether it remains within ratings under full power conditions.

Dave

dcgillespie wrote: Let's short the output terminals. Now we're pulling crazy current, with no load to absorb it -- except the plate! That sucker will be glowing cherry red, even though the bias is properly set. So while the correct amount of bias keeps the Pd in check under quiescent conditions, the load placed on the amplifier has by far the largest effect on whether it remains within ratings under full power conditions.

Dave

That's a concept I think I had a pretty good grasp of...

But that's brings up a question of how one would know if their speakers were creating excess load for certain types of high demand musical passages assuming proper load at the secondary, if someone was to bias the amp for max "sweat pouring from forehead" power. See if the tubes will cook a marshmallow while playing "Iron Man!"

Thanks much!