Bias question

.

[/quote]
(2 and 3)  
The feedback circuits are altered. Unless one can verify it's effectiveness with scope
and function generator this is suspect. Lacking these resources the best you can do is
restore to original configuration.
[/quote]

I understand that is the prevailing opinion.  For now ill keep it as is unless I find something particularly dangerous or unsafe.

The basic question is why is the bias so low and can't  be raised to what one expects it should be.
I suspect the answer is on the driver board.
My board has fewer resistors and more caps than oem.
Is there a clear wiring diagram for a pc3?

Thank you Bob for responding to this thread. Ill keep your comments in mind when I get to reading that article. If I decide to rebuild I will definitely purchase one of your boards. Hopefully there's something written down somewhere on this one.

The driver schematic turned out not to as hard as I was expecting. I imagine staring at the amp for 5 days has helped.

Much of the oem circuitry is gone.
Pin 9 no longer has a cap and resistor to ground
Feed back is only from the secondary of the OT.

Missing resistors.  330k, 10ohm x2,  18k x2,  270k x 2

Changed component values
270k resistor changed to 110k at pin 2
1.5M changed to a trim pot.  pin 3
.05uF changed to 1.0
620 ohm changed to 200
47k resistors changed to 30k and the one at pin 8 has an additional trim pot.
Coupling caps are .22
The rca has a 75k resistor to ground in stead of 470k

.

I agree the mods were done by someone with a plan.  

Looking back at the PS I have an explanation for the low voltage at C measured 350 should be 415. The relay has one set of contacts that shorts the 2k resistor coming off the rectifier. If the relay activates then 430v is applied to the buss.
The relay is a dpdt and has a latching circuit.
The rectifier socket is what triggers the relay. A different tube must be required here for the pins to be functional.
Any idea on how to figure out what tube is supposed to be there?
Maybe this is the cause of the low bias.

This was an easy test. Energize the coil on the relay by shorting pins 5 and 7. Then check dc voltages. Now they are running a bit high.
C 500
B 460
A 430
I can now set the bias to 1.5volts.
Which now leads to the next question.
The 4 trim pots on the driver board how do I go about setting them?

The one at pin 3.
is it acting as a volume control or right to left channel ballance?

The one at pin 8.
I'm guessing the voltage drop at pin 8 to ground should be balanced with the voltage across the 30 k resistor at pin 1?

The plate and cathode resistors must be slightly different values and that pot is for balancing the splitter.
The pot at pin 3 did adjust the voltage at pin 2.
I set it to 110v. being the voltage at B is a little over 400.
I'm going to go ahead and hook it back up see how it sounds.

Weather or not if its the right way to do it the pot at the cathode definitely adjusts the balance. I had two volt meters hooked across those resistors.

A bias resistor let go after 10 min with a signal.
I see what is meant by red plating.
Luckly I had a way to make 35 ohms so I'm back in business.
With higher buss voltages the lows have more authority,
The highs are as sweet as ever. And the break up when I lean on the volume is gone.
Let me put some hours on it and ill report back

peterh wrote:

Skelt wrote:Weather or not if its the right way to do it the pot at the cathode definitely adjusts the balance. I had two volt meters hooked across those resistors.

If the resistors are equal and the power tube is healthy and matched then there
is no place for adjustment.

I understand what you are saying and it makes perfect sense but I'm not going to change anything in the circuit for now.

Next up Emerson Lake and Powell

If tubes are running a little hot now , I would install 100e 10W resistor at each power tubes cathode and bias to 4V .

4V/100e = 40mA ; 40mA x 500V = 20W dissipation ( that's excluding power dissipated by G2 )

So can , in this scenario be biased to 3.8V .

First of all, the ST-70 was never built in a pentode configuration, it was ultralinear. It could be converted to pentode or triode operation, but pentode was rarely used as ultralinear gave better performance at a very slight decrease in power output from pentode mode. That said, if the green or green / white leads from the transformer are connected to pin 4 of the EL-34 output tube socket, then your amp is in ultralinear mode. I would advise adding a resistor on these leads if one is not present, 100 ohm to 1200 ohm 5 watt (non-inductive) are values to consider. 100 Ohm is the traditional value but others suggest a higher value that will decrease power output slightly but improve the sound quality. I use a 680 ohm in my ST-70.  If there is a jumper or a resistor from Pin 3 to Pin 4 (again, 100 to 1200 ohm, non-inductive), and the green and green / white leads are not connected to anything, then your amp is in triode mode.  

Personally, with your speakers ( I believe I saw they were Klipsch somewhere in the postings ), I would try the triode mode. You should still have plenty of power.

Regarding biasing, I would recommend 35 ma to 40 ma, not only will this improve tube life, but it will give a richer sound.

-Ed

One, add an NTC to one of the power cord leads in the amp. A NTC is a Negative Temperature Coefficient resistor that starts as a high resistance and drops to almost zero ohms as it warms up.

Why ?

One, your amp was built for 115 to 117 VAC, today it's commonly 122 or higher. The NTC will knock a couple of volts off that number and bring you tube heaters closer to 6.3 vac, which will extend their life. I believe a CL-80 is the proper value for an ST-70 ( 3 amp ), but some of my friends have added an CL80 to both the hot and the return leads to lower the voltage further or paralleled a pair of 2 amp CL-90s to get even a higher initial resistance. Measure the AC across EL-34 pins 2 & 7 to see what your allegedly 6.3 VAC actually is.

More importantly, the slow turn on will help protect you tube heaters from an inrush of current when you "flip the switch". Robert Tomer, a tube expert from the 1950s indicated in his book that an NTC will reduce the probability of tube heater failure by a factor of 4.

Second, if you have an accurate voltmeter with a 200 mv scale, consider the following; use a 1 ohm bias resistor between pins 1 & 8 to ground. Using the 200 mv scale on your meter, you read the bias current in equivalent milliamps, i.e. 100 mv measured = 100 ma. An additional benefit is lower overall amp impedance, your dynamics will probably be better.

I buy a 1 ohm, 1/2 % resistor from Mouser which gives me a precise measurement with my accurate meter.  

-Ed

I read the Van Alstine article and I agree with Bob. I believe adding a band pass filter on the input would not help to th transparency. I don't see any thing on Van Alstine schematics that matches with what I have.


I've put over 20 hours on it now since the resistor was replaced. Its not getting any hotter than the unmodded one. I'm using one amp per channel and a 10ohm resistor across the unused OTs.
The bias is holding steady at 1.5v.  I'm satisfied its holding together so it can come back to the bench for more analysis.

I can see now courtesy of my drawings that this  amp isn't in triode mode. I might try that later after further circuit discussion.

Here is an updated PS schematic. I added in the relay and fixed an error on the 1200uF caps. They are in series not parallel yielding 600uF and 500 volt ratings.
The voltages on the sketch are before and after the relay kicks in.
They are higher than what the dyna manual says they should be.
Are they too high for the rest of the circuitry?
A 420
B 460
C 500

Is there any benefit or harm with A and B in parallel? Both are hooked to C and aren't in series.

Solder Slinger wrote:One, add an NTC to one of the power cord leads in the amp. A NTC is a Negative Temperature Coefficient resistor that starts as a high resistance and drops to almost zero ohms as it warms up.

Good idea, Ill look onto putting one in especially since it had a SS rectifier.


Second, if you have an accurate voltmeter with a 200 mv scale

I do