Pitiful X-202

[ssouci]

I would go with the factory setting of 30 mA per tube. In my own unit, as well as other X-202 and SA-100 amplifiers I've worked with, the lowest distortion is produced at that setting (well, closer to 31 mA on average, but close enough). The Russian tubes should be able to handle it just fine. The optimum bias setting is only marginally a function of the particular tubes chosen (by brand, manufacturer, etc.), and virtually entirely a product of the circuit design they operate in. Of course, various tubes will all need their own particular setting to achieve that target, but the target itself always remains the same for a given circuit design and tube type, unless it is just an unreasonable operating point to begin with -- and there's a few of those units out there!
Dave

Q1: since it's so easy for me to set the bias, shouldn't I then use 31 mA for optimum performance?

Q2: I've noticed I can set the balance between tubes to precisely 0.00 and it's actually very stable... next I set the bias level, using just one of the tubes (I suppose because the assumption is that they will be the same since they've been balanced)... however, when I take a reading from the other tube out of curiosity, it's usually not exactly the same... so shouldn't I then fine tune the balance until they are?

I can witness the same thing using my dual "cathode current bias testing tool", which measures the current directly in each tube while running... when I switch back & forth between tube A and B in a pair (using the tool), the current is usually a bit different between tubes, eg. 30 and 31 mA... again, it would be easy for me to perfectly fine tune the balance (and absolute bias level) using this tool, so should I?

or am I trying to be too precise here with all this???

thanks!

 

[nj pheonix]

PROBABLY
All kidding aside. perfect is great! I think we've all fallen into precisely hand matching resistors, caps, voltages 3 places out. None of the vintage companies expected users to have more accurate tools than they had when they designed them. I've done it. I doubt at the point you're going to, even testing with distortion analyzer would net you anything, certainly not sonically . You're there. Sit back and enjoy the music!

 

[ssouci]

a few photos of my X-202...

 

[dcgillespie]

The socket adapters to monitor cathode current do exactly the same thing that the internal cathode resistors do. You can use either approach, your choice. Any difference you note between the two approaches is due to the tolerance of the resistors used both internally, and in the adapters. At the level you are balancing to, slight discrepancy is insignificant.

Dave

 

[ssouci]

help!... my bias readings are suddenly weird and WAY too high

my Fisher X-202 amp has been running happily for weeks now and sounds great... I had the output tube bias set to 30 mV, and was able to balance the tubes perfectly through the independent bias modification

the bias has been very stable over time and I could measure after 2 weeks, and it was still the same... plus once it stabilized during a measuring session, it barely fluctuated

BUT, today I wanted to measure some new tubes I got, wanting to use the Fisher amp as a "measuring device" to help me match my new tubes

so I swapped in some new tubes, and tried to set the DC balance and bias, but my bias readings were very high and came in at 60 and 80 mV at my lowest possible setting, which is way too high!

and this is despite having the Bias pot turned fully to minimum (ie., fully CCW)... turning them at all CW will easily drive the readings up over 100 mV

so I figured something must be wrong with the new tubes, so I yanked them out and put my old tubes back in, in their previous locations

but now those tubes are also reading way too high and I cannot begin to get them back down anywhere near 30 mV

any ideas???

- I did not change anything else with the amp from yesterday, and am using the same multimeter
- I have two 8 ohm dummy loads in lieu of speakers, and the wall voltage is only a wee bit higher then normal (was 119/120 V today, versus normal 118 V)
- all the tubes in question are the same NOS 6P14P-ER Russian tubes or the new Sovtek EL84M equivalent
- I have no inputs, and Volume is at minimum, and selector to Aux

thank you

 

[ssouci]

update: I just tried the take bias readings instead using my "bias testing tool" that fits into the tube sockets and plugs into my multimeter

with bias pot at minimum position, I got approx. 7 mV... and then by turning the pot CW, was able to get the readings up to 30 mV.... so this method seems to be working normally!

and it seems that taking the readings from the rear speaker terminals (where I had installed the independent bias modification) is no longer working correctly... I tried there again and the readings from there are still very high, and on one channel even "out of range"

wtf???

 

[dcgillespie]

If everything is connected correctly, that could only happen if the 10 Ohm resistors you installed are no longer 10 Ohms, but now much higher.

Dave

 

[ssouci]

If everything is connected correctly, that could only happen if the 10 Ohm resistors you installed are no longer 10 Ohms, but now much higher.

Dave

OK, I will open her up and check these for correct resistance

if that's the issue, any idea of what might have caused this?... all I did was put in a set of Sovetek EL84M tubes, which are apparently virtually identical to old 6P14P-ER (the tubes that were running fine)

 

[dcgillespie]

If the test points are connected on the back connector strip, then with the unit turned off, you should be able to just measure from each test point to ground with your Ohm meter to check the resistors. No need to open it up to check them! About the only thing that can damage the cathode resistors is in fact the tubes themselves, but you'll usually hear a distinct "pop" in the speaker when they blow. If you didn't, but yet your bias readings were really wonky, just remember the one thing that can always get in the way of a good diagnosis: Something you think is true......isn't.

Dave

 

[ssouci]

p

If the test points are connected on the back connector strip, then with the unit turned off, you should be able to just measure from each test point to ground with your Ohm meter to check the resistors. No need to open it up to check them! About the only thing that can damage the cathode resistors is in fact the tubes themselves, but you'll usually hear a distinct "pop" in the speaker when they blow. If you didn't, but yet your bias readings were really wonky, just remember the one thing that can always get in the way of a good diagnosis: Something you think is true......isn't.

Dave

OK, I just measured each resistor as exactly 10 Ohms (I used precision resistors)... also, I opened the chassis and inspected... these and all other resistors in the output tube area appear to be intact with no discoloration (I recently replaced every resistor here with new metal films)... so I guess that all the resistors are perfectly fine... seems very mysterious and I will just have to try again

thanks!

 

[ssouci]

OK, problem solved!... it was an issue with my multimeter (!!!!!) and has been corrected... so the bias setting system is all good again, thanks Dave!!!

but now I have different (possible) problem...

as I've been testing the amp, when I happen to brush my hand or finger along the brass faceplate, I feel a weird "vibrating" sensation, that I'm guessing means there must be some sort of electrical charge on it... I do not feel any shock, only the vibration as I rub

it is easy to replicate the issue and the vibration was always there... however, shutting off the power causes it to immediately disappear... switching the plug polarity makes no real difference

the amp is sitting on a piece of cardboard, on a carpet overtop a hardwood floor... the bottom cover is on and there are 4 rubber feet installed... I am using a 2-prong plug (probably original that appears to be in very good shape)

what the heck could be going on, and is this safe?

thanks!

 

[fred soop]

Classic case of leakage. Measure between chassis and a good ground. Without the test network described below, you may get some high voltages, especially with a digital meter. However, there is probably enough resistance to limit the current to safe levels.

Standard leakage test copied from one of my earlier posts .....

The high chassis to ground voltages may be the result of using a very sensitive digital meter. The standard leakage test is a 1500 ohm resistor in parallel with a 0.15 µF capacitor, this combination connected between chassis and ground.

The equipment under test is to have no other connections except power. No antenna, no cables to other equipment, etc.

The test is performed with all combinations consisting of both orientations of the wall plug and power on and off.

Measured voltage across the RC network should not be more than 0.35 volts for any of the above combinations. If more than that, leakage is considered excessive.

The line capacitors should be replaced with proper XY capacitors. X is line to line and Y is line to chassis.

Digi-Key BC2379-ND, Vishay VY1472M63Y5UQ63V0, at 71 cents is X1Y1 rated (highest rating for XY capacitors) and is suitable for all X and Y applications.

Even if all capacitors are removed from the line, there will still be capacitive coupling within the power transformer.

 

[ssouci]

UPDATE: when I hook up the amp to my speakers and a couple of inputs (CD player and PVR), the vibration seems to disappear!... that's probably why I haven't noticed it before

does this provide any clues?

 

[dcgillespie]

The power transformers in the X-101ST and X-202 are famous for having a little leakage to produce the sensation you are noticing. You can usually fix it by simply reversing the plug in the socket. The X-202 does not use any AC line bypass caps.

Dave

 

[ssouci]

reversing plug didn't seem too make much difference, however plugging in the speakers and CD player and a PVR did, and I no longer feel anything

is this dangerous, or any reduction of SQ?

thanks!

 

[ssouci]

bias setting fine-tuning question:

when I set the bias (using my independent test points, now located externally on the speaker terminal strip), my normal method for each tube pair is to:

- first dial back the Bias pot fully CCW to lower the bias to minimum
- set the DC Balance by adjusting for 0.00 mV DC between test points (of each tube pair)
- increase bias by turning the Bias pot CW, until it reaches 31 mV (as measured between the test point and chassis)
- do the same for the other channel
- go back to the first channel and check/adjust bias again, because it always changes after the other channel gets adjusted
- continue with a few back & forth readings/adjustments until both channels are equal

BUT, I've noticed that the bias reading will be different, *depending on which tube* you happen to measure from (despite the DC balance being initially set to zero)... for example, the bias could be 30.4 mV from one tube, and 31.0 mV from the other tube

so which is the correct tube to use?... and why don't I just tweak the DC balance pot a wee bit so that both tubes are the exact same?... wouldn't that leave me with perfect BC balance and bias across all tubes?

I know I'm probably getting way too picky here, but my bias is so easy to set and so stable that I'm able to set these dead on with no extra work, so why not?

thanks!

 

[larryderouin]

Just balance them across the pair and be done with it. It's a circular problem, as the adjustments are interactive. Any little adjustment on one will affect the other one of the pair. Get them balanced, both pair. Then check the bias. Adjust one pair or both to get to your target. Check and tweak each pair's balance. If tweaking is NOT indicated, close it up and put the tools away. It will STILL SOUND GOOD with a 3-4ma imbalance. OCD can be overdone at times. Humans are inherently IMPERFECT. The "PERFECT BALANCE AND BIAS" can and WILL drift over time(6 months). And you won't hear it either. None of the 4 will track exactly the same. All you can do is get them close and enjoy them.

 

[dcgillespie]

The goal, and correct final result, is to have a 0.00 vdc (or reasonably close) differential between the two output tubes of each channel, with them each drawing the correct quiescent current of 30 ma per tube. The fact that you balanced the tubes at a very low current draw, and then find them unbalanced again as you bring them up close to the correct quiescent current draw is due to the characteristics of the two tubes not being perfectly matched at every point along the grid curve. Even well matched tubes won't be absolutely identical in this regard, but minimizing the discrepancies between them always minimizes distortion. This is a perfect example showing why I have always advocated that just because a unit offers a full set of bias and balancing controls, you still want to use matched tubes in it anyway. The closer the tubes are inherently balanced without the need for circuit correction to obtain that balance (which is what balance controls do), the better the circuit will perform, and the ultimate performance will be.

Dave

 

[ssouci]

Just balance them across the pair and be done with it. It's a circular problem, as the adjustments are interactive. Any little adjustment on one will affect the other one of the pair. Get them balanced, both pair. Then check the bias. Adjust one pair or both to get to your target. Check and tweak each pair's balance. If tweaking is NOT indicated, close it up and put the tools away. It will STILL SOUND GOOD with a 3-4ma imbalance. OCD can be overdone at times. Humans are inherently IMPERFECT. The "PERFECT BALANCE AND BIAS" can and WILL drift over time(6 months). And you won't hear it either. None of the 4 will track exactly the same. All you can do is get them close and enjoy them.

thanks Larry... it does drive you a bit crazy how these things interact... nevertheless, I'm still trying to do my best just to get it as close to perfect as long as it doesn't take too long, mainly make myself feel better I guess, and hopefully also for the benefit of the amp

 

[ssouci]

The goal, and correct final result, is to have a 0.00 vdc (or reasonably close) differential between the two output tubes of each channel, with them each drawing the correct quiescent current of 30 ma per tube. The fact that you balanced the tubes at a very low current draw, and then find them unbalanced again as you bring them up close to the correct quiescent current draw is due to the characteristics of the two tubes not being perfectly matched at every point along the grid curve. Even well matched tubes won't be absolutely identical in this regard, but minimizing the discrepancies between them always minimizes distortion. This is a perfect example showing why I have always advocated that just because a unit offers a full set of bias and balancing controls, you still want to use matched tubes in it anyway. The closer the tubes are inherently balanced without the need for circuit correction to obtain that balance (which is what balance controls do), the better the circuit will perform, and the ultimate performance will be.

Dave

thanks Dave! (once again...)

Q1: so wouldn't it be better to concentrate on setting the DC Balance up in the 30 ma range? (and perhaps still doing so first at low current draw, just to get things in the ballpark)

Q2: is the ultimate goal and function of the DC Balance and Bias controls to get every bias reading (from all 4 test points) to read exactly 30 ma?

Q3: if that's achieved, will the DC balance then automatically also be zero (and if I put my meter across test points for a tube pair it will read zero volts)?... sorry, but I'm a bit unclear if the DC balance and bias readings are "coming from the same source" and "interlocked", or if they are slightly different things