Hiwatt DR103 voltage problem on new build

[dvlp]

Hi all. I have a voltage problem on a new build. This is my first build. I used Mark Huss's diagrams and schematics (thanks Mark). The only changes are I used a PT with a 70v bias winding and I added 2 bias adjustment pots.

With the tubes out I measure 460v at the plate pin #3. With the bias voltage set to -35v I measure a 15mA load on the secondary. The secondary is rated 300mA @ 345vac.

The problem is when I install the tubes the plate voltage drops to 334v, the bias voltage drops to -27v and the load on the secondary goes to 300mA. If I adjust the bias voltage lower (to raise the tube bias) the load on the secondary climbs quickly past it's rated 300mA and the plate voltage drops proportionately.

With tubes in the 1st cap after the rectifier charges to 220v with standby off then drops to 170v with standby by on (power to other caps).

I've tried different tubes. I tried un-soldering the bias pots and wiring the bias circuit to the schematic and no change. I double checked my work to the diagram and checked the diagram against the schematic the best I could focusing mainly on the power and output stage. I double-checked the capacitors to make sure they were wired to the schematic. The only thing I see odd is with the amp turned off all the capacitors measure to spec except the 50 + 50 uF dual. It measures 86uF on both sides. If I un-solder the leads linking it to the pre-amp section the reading drops back to normal. From probing with the DMM it looks like the 32+16 uF dual cap in the pre-amp section is affecting the reading. This might be normal, I don't know.

If someone has ideas for things to check I sure would appreciate it.

[catalin gramada]

Draw here how your bias supply looks like,please.
The bias voltage means nothing in this moment. You should be able to know the specific current for each power tubes at iddle in respect with plate voltage you get. Have care to not exceed more than 70 percent power dissipation from max value of you power tube for you specific plate voltage and OT primary winding load. Instal few ohms
precise resistors in power cathodes to found exactly how much current each tube draw.

[Mick Bailey]

The capacitor value is not an issue in itself (caps often read over their nominal value), and any capacitor measured in-circuit where there are parallel components can give a misleading reading.

My initial thoughts on excessive current draw when tubes are inserted is oscillation. Firstly, do you have the speaker or dummy load connected? This is essential. Do you have a scope?

[Dave H]

With the tubes out I measure 460v at the plate pin #3. With the bias voltage set to -35v I measure a 15mA load on the secondary. The secondary is rated 300mA @ 345vac.

The problem is when I install the tubes the plate voltage drops to 334v, the bias voltage drops to -27v and the load on the secondary goes to 300mA.

If someone has ideas for things to check I sure would appreciate it.

I don't think you have enough bias voltage. Turn the bias voltage to its maximum setting (I'd aim for -50V) before inserting the tubes.

[Chuck H]

What they said^^^^^^^^^^^^^^^^
And there is also the possibility of a miswire at the tube sockets causing this excessive current. As suggested, I'd start by going to a much higher bias voltage (if more negative voltage can be called higher?) and see what happens. If you still have anomalous readings, but the amp is stable enough to measure some things without melting down then some voltage readings from all the power tube pins might help to narrow down the problem.

[Dave H]

The secondary is rated 300mA @ 345vac.

If the above is for the B+ winding it could be the reason your voltages are sagging. 300mA isn't enough for a 4 x EL34 100W amp.

[Mick Bailey]

It should bias up OK - only needs about 120-150mA to sit there idling. The voltage shouldn't drop to the degree it does if the bias voltage is correct.

[Dave H]

It should bias up OK - only needs about 120-150mA to sit there idling.

Yes, I checked after I posted and Mark Huss says the original Partridge DR103 power transformer was 350V @ 300mA. That's only about 100VA which doesn't seem enough to me for a 100W amp (unless it's 100% efficient )

[dvlp]

Thanks for the help guys

Catalin - I think you're asking for a diagram and not a schematic. I can try to create a schematic if that's what you want. I originally wired in the bias pots as shown in the attached image but to eliminate that circuit from being the cause I un-soldered the connections and wired in the original bias supply layout. The only difference from the original is I changed the 47k bias range resistor to 10k. The voltage sag was still present so I re-installed the adjustable pots.

I have 1 ohm 1/2 watt resistors from pin #8 (cathode) to ground to convert mA to mV readings. I think that's what you're asking.

Mick - I'm using an 8" speaker rated 200 watts at 8 ohms as a dummy load. After your suggestion I tried plugging the amp into a 2-12", 8 ohm cabinet and no improvement. I don't have a scope. The only tools I have are a Fluke 87V and a Variac.

Dave -
-50v bias
432.7v plate
4mV tube bias

-40v bias
399 plate
13mV tube bias

-30v bias
348 plate
25mV tube bias

Chuck - 2 readings
pins 1&8 cathode 4mV across 1 ohm resistor
pin 2 heater 3.162 a/c
pin 3 plate 432.2v
pin 4 screen grid 431.4v
pin 5 grid control -49.03v
pin 6 not used
pin 7 heater 3.166 a/c
-------------------

pins 1&8 cathode 13mV across 1 ohm resistor
pin 2 heater 3.017 a/c
pin 3 plate 400.3v
pin 4 screen grid 400.6v
pin 5 grid control -39.61v
pin 6 not used
pin 7 heater 3.015 a/c

[Chuck H]

I hope there's an error in your pin designation because as it's written you have the screen supply on the control grids and bias voltage on the screen grids.

[g1]

I'm not sure what the problem is. As you make the bias more negative, the tube idle current decreases and the plate voltage increases.
This is what your numbers correctly show in post #9 (aside from the discrepancy Chuck noted).
What are you trying to fix?

[catalin gramada]

I hope there's an error in your pin designation because as it's written you have the screen supply on the control grids and bias voltage on the screen grids.

[dvlp]

I hope there's an error in your pin designation ...

That was a typo. I corrected post 9

[catalin gramada]

Are you sure read corectly 4mV over 1 ohm resistor? It is not 10 ohm ...maybe? I remarked you heater voltages going down aso that means seriously stress on power transformer not something from 4 to 13 mA (16 to 52). So Where this current gone ?

[dvlp]

I'm not sure what the problem is.

The plate voltage drops dramatically and the secondary reaches it's max rating before I am anywhere near a reasonable tube bias current. I can't really bias it with such wild voltage swing. If I try to bias the el34s at 70% for 440v plate I would need to see about 40mV/mA at the cathode. By the time I get to that target the plate voltage has dropped more than 100v and the secondary load is way beyond it's rating.