DRRI bias sweep.

[Sea Chief]

Sorry I need to start completely afresh.

1) What is the higher negative voltage of these two: -33v / -25v ?

2) Do we discuss it oppositely to the way it a) actually is and/ or b) how its technically referred to as?

3) Am I after a higher or lower negative voltage.

4) why?

thanks SC

[martin manning]

Think of it like an outdoor thermometer. Which is the higher temperature, -25 or -33? Answer: -25 is 8 degrees hotter than -33. Same for voltages.

The higher the bias voltage is, the higher the idle current is. -25V is higher than -33V, and will allow more idle current than -33V. -25V is "hotter" bias than -33V, same as the thermometer.

Idle current is set such that the idle plate dissipation, which is the wattage that the tube has to dispose of when there is no signal present, is within safe limits. 6V6's are nominally capable of dissipating 12W from their plates, and a reasonable place to start is setting the bias such that they at 60-70% of their limit.

You calculate that like this:

For 70% dissipation, 12W * 0.7 = 8.4W. Power in W is V * A. Suppose your plate voltage is 400V. The plate current should then be set to 8.4W/400V = 0.021A or 21 mA. If you are measuring the current at the cathode using a current sensing resistor, you have the screen current (which is about 10% of the total) mixed in. That means your plate dissipation will really be about 10% lower than 70%, or 63%. Make it easy on yourself and use 1 ohm resistors between the cathodes and ground so the voltage you measure is numerically equal to the current in amps.

[Sea Chief]

Ok hi Martin. Im getting a * symbol where something should be so i cant make too much sense of "the Power in W is V * A".

But what say I have a plate V of 413v, and I have screen resistor (no not ideal but there we go) of 1.6r Im measuring across. What am I aiming for, to get 70% dissipation, on the plate current (& is "plate current" mean the same as the 19mA bias figure I am reading now?).

thx SC

[martin manning]

The "*" is for multiplication. Volts times amps is watts.

Get your terminology straight or you won't be able to communicate. For 6V6:

The plate (or anode a) is pin 3, which connects to the OT primary.
The screen grid (g2) is Pin 4, which will be connected to the power supply through a 470 ohm resistor (typical for a DR). Most people will just call this pin "the screen."
The control grid (g1) is pin 5, which is where the signal is applied. Most people will just call this pin "the grid."
The cathode (k) is pin 8, which is either grounded directly or grounded through a current sensing resistor so you can measure cathode current.

If your plate voltage is 413V, then you want (for 70% dissipation) 8.4W/413V = 0.020A = 20mA. If you have a 1.6 ohm resistor between the cathode (pin 8) and ground to measure cathode current, then the voltage across that resistor should be set to (ohm's law) V = I * R, or V =0.020 * 1.6 = 0.032V or 32 mV.

How do you know this current sensing resistor is 1.6 ohms? is it marked 1.6 ohms, and is it a 1% tolerance part?

[Sea Chief]

Ok understood that Martin thanks. No of course I dont know if its a 1% telerance part - I get our point. Why was I advised to put this in if its such a wayward way of measuring-?

Ok I will sand an area of chassis next to each pin 8 and solder this R to ground, then measure directly off pin 8. I think thats the idea?

Then next thing is to get the bias pot to sweep much more so I can reach this new 20mA figure. Leo is saying I need to -decrease- the piggybacked 22K (onto the existing DR 22k) to get the sweep wider. This though opposes what I did: the orig GZ34 had the 22k. So with my 5U4 I put in, dropping a big V to 430v b+ (as I was 480V with the gz34) I added the extra 22k as the bias pot sweep was to slim. Surely if Ive dropped the V further with my 5R4 (413V b+) I need to increace this R further... nit decreace it as Leo is suggesting-?

I had no idea the 20mV reading needs to be divided by 1.6 to get the actual figure of my mA bias at pin 8. Am I thinking tight here?

[martin manning]

How are the 6V6 cathodes grounded now, and where do you have the 1.6R resistors? How do you know the resistors are 1.6 ohms? If you are measuring them, I would not trust the reading unless you have something better than a typical DMM. You are correct that the current sensing resistor should be connected between the cathode (pin 8) and ground, and the voltage to be measured is on pin 8.

Measuring cathode current this way is a very reliable and safe method, but I strongly recommend you get some 1 ohm 1% resistors to be sure that you are getting a good reading.

Reduced plate voltage will require increasing the bias voltage (making it less negative), and higher idle current to get the idle dissipation where you want it. For 8.4W dissipation:

8.4W/450V = 0.019A

8.4W/400V = 0.021A

Reducing the value of the 22k resistor (R59, I believe) by placing another resistor across it would be the right move, since that will bring the available bias voltage range up closer to zero (ground).

[martin manning]

I had no idea the 20mV reading needs to be divided by 1.6 to get the actual figure of my mA bias at pin 8. Am I thinking tight here?

Yes correct. I = V/R.

[Sea Chief]

The 1.6r's are on pin 8. I just solderd in a 1.6 r between the pin 8 wire and the pin 8 terminal, then I measure across in mV and that can be read in mA.

I had no idea I was meant to multiply b 1.6 for a proper reading of the mA.

At the moment I have two 22k in piggyback at R59 (I added the 2nd, as explained bc I replaced the orig gz34 for a 5U4 as I had huge 480v b+). And I only read 19mV across the 1.6r (or divide that now by 1.6 = 11mV max on the pot then?!).

What value should I replace the piggybacked 2nd 22k with instead?

[martin manning]

The 1.6r's are on pin 8. I just solderd in a 1.6 r between the pin 8 wire and the pin 8 terminal, then I measure across in mV and that can be read in mA.

That should be fine, no need to solder to the chassis. Just verify that you get zero volts on the ground end of the current sense resistor.

I had no idea I was meant to multiply b 1.6 for a proper reading of the mA.

To be clear, once you calculate the mA you want for 70% dissipation, you then multiply that value by the current sensing resistor value to get the mV value you want to see on pin 8. With 413V on the plates, and 1.6R, you are looking for about 32 mV.

At the moment I have two 22k in piggyback at R59 (I added the 2nd, as explained bc I replaced the orig gz34 for a 5U4 as I had huge 480v b+). And I only read 19mV across the 1.6r (or divide that now by 1.6 = 11mV max on the pot then?!).

What value should I replace the piggybacked 2nd 22k with instead?

The two resistors that determine the bias range are R69 (22 ohms), and R59 (10k). With a 22k across 22 ohms you have done almost nothing for the bias range, and you would need to increase that 22R to get a higher bias voltage. Is that what you have now, a 22k across a 22 ohm?

The other way to go is paralleling another resistor across R59 to make it smaller. I'd try a 10K, but I suggest that you pull the power tubes out and measure the available bias range before you do that, and again afterward.

[martin manning]

Sorry, I see you said you have two 22k at R59. Does that mean there are three resistors there, two 22k and the original 10k?

[Sea Chief]

I remember the R69 as being part of the bias circuit, bc the chap looking over my amp (a pro) just for a once-over, really due to it sounding very flat I thought as it is, acidentally shorted 2 lugs on the bias pot = a fried R69. I redid this myself a few days later, but the red-plating of the 6v6's (we saw & couldnt work out why/ but diagnosed within 20 mins or so, thank heck)
alarmed us.

Just to be clear. I have only jumpered another 22k onto the back of the 22k (R59: says 10k on the shematic, but I found it was actually 22k/ maybe an upgrade since the 1st schematic perhaps). So these two are in parallel.

In order to get 32mA, do I therefore have to 1st increace the negative V to maybe -25V and then I see an increace in the mA on the bias pot sweep.. of so how do I do this, and why then does this procedure seemingly not include the changing of R59?

or does changing R59 mean that the negative V increaces as a result (IE is that the whole idea of changing R59?).

Thanks SC.

[Sea Chief]

I thought that Leo siad I didnt want to see anything below -30V, a fig of -35V was ideal or so. So why then would I want to dleiberatley drop the neg V to below -30, for say -25V.

Am i thinking updside-down or something?

Christ this is complicated..

[martin manning]

So there is 2x 22k = 11k in R59 now, and what is in R69, where the 22 ohm that burned was?

[Sea Chief]

I thought that Leo siad I didnt want to see anything below -30V, a fig of -35V was ideal or so. So why then would I want to dleiberatley drop the neg V to below -30, for say -25V.

Am i thinking updside-down or something?

Christ this is complicated..

Where has Leo gone?? all his posts have been deleted. I was using them for info, terrific help you were as usual I was hugely grateful. Have I done s'thing to upset him?

[Sea Chief]

So there is 2x 22k = 11k in R59 now, and what is in R69, where the 22 ohm that burned was?

22R back in R69 (fried).

Yup 2x 22k in R59 (shematic says 10k for some reason).