Taking Readings From The OT To Use The Shunt Bias Method

[Gary_S]

What's the best way to go about this? I've read of amp builders like Ken Fischer and Gerald Weber who say the shunt method is the best way of setting an amp's bias but you need to take a reading from the OT for this.

The OT is bolted on Marshall combo's to the chassis and so you'd need to undo all the bolts to gain access to the insides, then you have to take a reading with it loose like that? or would you reattach it to the chassis but with the plate covers off?

It seems an awkward way to set bias. I just use a bias probe and set it that way but was wondering what the rest of you guys do? do any of you use this shunt method to bias your amps? Is it better than just using the probe?

[six el six]

Insert an ammeter in series with your OT center tap. You don't unbolt anything.

There's a bunch of ways to see how much idle current your tubes draw and that's not a method for someone who doesn't really know wht they're doing.

I think I should tell you to please be careful. Getting shocked by a live tube amp, at a minimum, really really sucks.

I'm telling you that because your question leads me to think you might need some one on one, in person direction.

A bias probe would be much safer. Difficult to kill yourself (or your cat) with a bias probe.

[PRR]

> you'd need to undo all the bolts

PLEASE just put 1-Ohm resistors in the CATHODE leads and measure that way.

(Yes, this includes screen current but that's no big deal.)

As El Six says, you don't unbolt, you never "go inside" an OT (unless it's headed for the trash). Ken and Gerald probe the OT _leads_, available at tube plate pins and power supply. However this exposes their hands to 440 Volts, and puts the same amount on all the insides of the meter. It also requires some information about OT resistance.

While 1-Ohm resistors do mean "modifying the amp", after that (usually simple) mod any future bias-checks are simple and SAFE.

Bias-Rite or similar DIY probes (kit available from this forum's host, "Tools" page) are another good way to go.

[tubenit]

PRR gave you excellent advice.

I have not & will never use the "shunt method". Not worth the risk to me.

With respect, Tubenit

[Gary_S]

Insert an ammeter in series with your OT center tap. You don't unbolt anything.

The two red leads going to the OT are the output from the plates and they attach to either end of the primary of the OT. So the other wire, i can't remember what color it is, it may be brown, that's the one that go's to the center tap of the transformer. How do i hook it in series with the OT? I'm not looking to do this i just want to understand it for future reference!

In Ken Fischer's writings he said connect one end of your meter to the CT of the OT and the other lead to the ouput plate of your power tube.

I just actually set the bias on my Marshall yesterday using a bias probe which was very simple but i want to understand the mechanics of the other way of doing it as well.

[HotBluePlates]

I've read of amp builders like Ken Fischer and Gerald Weber who say the shunt method is the best way of setting an amp's bias ...

It is not the best way, in my opinion. It does, however allow you to measure only the plate current, while inserting a 1Ω resistor in the cathode also includes screen current.

To do it, you use an insulated alligator clip attachment for your meter, and clip the black lead to the output tube plate. Clip the red lead to the B+, at the OT center-tap. Set your meter to read current (in most meters, you have to move a lead to a different socket). NOW, turn the amp's power on without your hands in it.

Turn the amp off, move the black lead to the other output tube plate(s).

If you try to do any part of this without clipping the leads in place, you WILL get arcing/sparking that will startle you, make noise, and let you know you're all primed for electrocution. I haven't even attempted OT shunt in >12 years because it is time consuming and dangerous compared to the 1Ω resistors. If you slip, you can connect B+ to pin 2 (filament) and toast your PT.

I just don't feel I gain any valuable knowledge with OT shunt that can't be gotten some other way. It's also possible to find meters that measure d.c. flowing through a wire without interrupting the wire or circuit.

[Gary_S]

Clip the red lead to the B+, at the OT center-tap.

Thanks HBP for your advice. See the bit above is what's confusing me; how do you clip to the center tap when you can't access it? are you clipping to a point in the circuit that leads to the OT center tap and measuring there?

If you try to do any part of this without clipping the leads in place, you WILL get arcing/sparking that will startle you, make noise, and let you know you're all primed for electrocution. I haven't even attempted OT shunt in >12 years because it is time consuming and dangerous compared to the 1Ω resistors. If you slip, you can connect B+ to pin 2 (filament) and toast your PT.

Ha! it sounds a bit hairy!  don't worry i won't be attempting this any time soon. As you know i just got the problem with my Marshall with the grid bias sorted out. So last night after fitting the new tubes i got my bias probe out and biased the amp for the new one's.

I was just reading through my books about the shunt method of biasing and i wanted to get my head around it and get it set in my mind exactly how they were doing it.

Thanks HBP.

[HotBluePlates]

Clip the red lead to the B+, at the OT center-tap.

Thanks HBP for your advice. See the bit above is what's confusing me; how do you clip to the center tap when you can't access it? are you clipping to a point in the circuit that leads to the OT center tap and measuring there?

If you have to ask, you should not do OT shunt!

I know that sounds like a jerk-answer, but I've been bit by 450vdc before, and you don't want that experience. I was very fortunate it went in through a fingertip and out the forearm.

So real answer for learning:
I don't open the transformer to get phycial access to the center-tap, and neither did Ken Fischer or Gerald Weber. You probe at a point that is electrically the same.

Look at the diagram below. The red line runs from the physical OT center-tap to its connected components. The OT CT runs to the first paired set of filter caps after the rectifier, and also the HT fuse and fuse-side of the choke. Probing at any of those points is the same as probing directly at the OT CT.

The OT CT is also probably a red wire running from the OT to one of these points. It has to make a connection to supply B+ voltage from the power supply through to the output tubes.

I was just reading through my books about the shunt method of biasing and i wanted to get my head around it and get it set in my mind exactly how they were doing it.

So here's how:

When  your meter measures voltage, it tries to act like a large resistor. Voltage can exist across that big resistor without any significant current being diverted through the parallel path represented by the meter.

When you switch your meter over to read current, it tries to act like a good-quality short-circuit. That's because now the meter wants current to flow through it, so the current can be measured.

If you read any book or manual about how to make electrical measurements, they'll tell you to measure current you need to break the circuit, and use the meter in series to complete the circuit. That forces circuit current to flow through the meter.

But the OT shunt method seems to be connecting the meter like a voltage reading. How can it measure current?

Because the meter resistance in current mode is very much lower than the winding resistance of half the OT. So current diverts away from the OT CT and through the meter, on its way to the output tube plate. Looked at like this, you'll see why the red lead needs to connect to the OT CT, and the black lead to the output tube plate. That's cause the plate is "closer to ground" that the OT CT, which is at B+ voltage.

Ha! it sounds a bit hairy!  

You don't know the half of it (literally).

The other half is what happens after you're done measuring your output tube current safely. I guarantee you will next go to make a voltage measurement, short something out and cause minor/major damage as things burn.

Why? Because you forgot to unplug that meter lead from the current socket, and put it back into the voltage socket before making your measurement. And you probably won't remember to turn the dial back over to "V" because you use the current setting so few times. So when you went to measure voltage, your meter was still set up for current and short-circuit the voltage you tried to measure.

Ask me how I know you'll make this mistake...  

[alerich]

(quoting author Gary S)

It seems an awkward way to set bias.

It is and it can also be dangerous to your and/or your amp if you don't do it correctly.

I just use a bias probe and set it that way but was wondering what the rest of you guys do?

I use the 1 ohm sense resistor method whenever possible... the bias probe when that is not feasible.

Do any of you use this shunt method to bias your amps?

Never have, never will. Too few benefits - too many risks.

Is it better than just using the probe?

Not really. It allows you to measure only the plate current but screen current is so small as to be ignored. Also, the addition of the screen current in the sense resistor/bias probe method adds a little margin of safety to your settings.

[Gary_S]

Ask me how I know you'll make this mistake...  

Thanks for the post HBP there was a lot of good info in there and i learnt a lot.

I couldn't find the diagram in the post to look at.

No i won't make that mistake, i hope! no i always check double check and triple check what i'm measuring before i do it. but there doesn't seem to be much reason to set the bias that way by all accounts.

Just to summarise then it's a matter of probing the lead from the OT where it joins the HT fuse, probe that with the red lead and touch the black probe to pin 3 the plate with the meter set for ma's current?

[tubenit]

Just to summarise then it's a matter of probing the lead from the OT where it joins the HT fuse, probe that with the red lead and touch the black probe to pin 3 the plate with the meter set for ma's current?

This sounds like you are considering doing something that almost every response stated was an unnecessary risk and/or dangerous?

Was there one response that suggested this is a good idea or a better way of measuring bias?

With respect, Tubenit

[Gary_S]

This sounds like you are considering doing something that almost every response stated was an unnecessary risk and/or dangerous?

I've already biased the amp tubenit. I don't need to bias it again. I said earlier in the thread i'd used the bias probe method.

I simply want to understand the process and why some techs think this is the best way to bias an amp.

If Ken Fischer and Gerald Weber think it's a good way to do it i want to know why.

Was there one response that suggested this is a good idea or a better way of measuring bias?

With respect, Tubenit

Not on here but if those two gentlemen think it's the best way why is that?

[sluckey]

If Ken Fischer and Gerald Weber think it's a good way to do it i want to know why.

Probably because it's more accurate (no error due to screen current). It's a quick, simple, easy, straightforward method that directly gives you the tube plate current with no calculator.

But, it's risky! Even in the hands of an experienced technician. Personal injury, damage to the amp or your test equipment can happen very quickly. Let your probe slip from pin 3 to pin 2 and damage will occur almost instantly.

I don't use the OT shunt method for all the above mentioned reasons.

[Gary_S]

If Ken Fischer and Gerald Weber think it's a good way to do it i want to know why.

Probably because it's more accurate (no error due to screen current). It's a quick, simple, easy, straightforward method that directly gives you the tube plate current with no calculator.

But, it's risky! Even in the hands of an experienced technician. Personal injury, damage to the amp or your test equipment can happen very quickly. Let your probe slip from pin 3 to pin 2 and damage will occur almost instantly.

I don't use the OT shunt method for all the above mentioned reasons.

Thanks sluckey, this clears it up for me.

I just wondered why those guys used the method and described it in their writings if it was an awkward and risky thing to do.

Thanks i appreciate it.

[HotBluePlates]

I just wondered why those guys used the method and described it in their writings if it was an awkward and risky thing to do.

I'd advise you to take a lot of what Gerald Weber says with a shaker full of salt.

While he does have a knack of explaining things such that a newcomer can understand it easily (and that once included me, too), some things he says are incompletely explained or are subjective. Gerald also was a salesman before venturing into amps, and he has a tendency to greatly hype things.

Secondly, you have to read their comments in context.

They originally made those recommendation in the early-mid 90's. If you'd gotten a copy of Aspen Pittman's Tube Amp Book (4th Edition), it would tell you to bias by applying a test signal to the amp and monitor a dummy load in place of the speaker with an oscilloscope. You then crank the amp up until the output shows visible distortion, then back the volume off a hair. You'd then tweak the bias adjustment until a notch appeared where the output signal crosses over the 0v point ("crossover distortion"). At that point you'd adjust for a little less bias voltage (resulting in more idle current), until the notch disappeared.

Their argument was that every time you tried to "bias with an o'scope" you'd land on a different result. Instead, they advocated for biasing by monitoring tube current, as your results could always be repeated.

Now, Ken Fischer had worked at Ampeg, and lots of their amps before and during his time there had 10Ω resistors in the cathode circuit. It was 10Ω instead of 1Ω because that allowed good resolution of the reading on typical meters readily available at the time. So 20mA = 200mV, landing at an accurate and easy-to-read point on an analog scale.

So we're saying, sure, bias by current... just not their way!

Anyway, a last reason the not to use OT shunt as they describe is suggested is something I alluded to earlier:

It's also possible to find meters that measure d.c. flowing through a wire without interrupting the wire or circuit.

I had in mind the Hewlett-Packard 428A. That's 1958 hi-tech, I'm sure there something similar from a modern company. This is a much safer method, if you just have to measure the d.c. flowing from the tube plate to the OT.

Throwing more mud in the mix, there are times it is perfectly valid to bias by measuring only the applied bias voltage. That doesn't account for variations in the tubes used, but does bias the amp to the operating point intended by the designer.

[PRR]

> touch the black probe to

No. Well, that works, but---

As HBP said (and you quoted): If you try to do any part of this without clipping the leads in place, you WILL get arcing/sparking that will startle you, make noise, and let you know you're all primed for electrocution.

When you 'touch', the tube current in the winding diverts to the "short cut" through the meter. When you un-touch, that current diverts back to the winding.

When you instantly start/stop current in a large winding, there is a huge voltage spike. As you pull the meter probe off, the voltage kicks, arcs back to the probe, arc breaks, kick again, re-arc. (If you've used an old-fashioned welder, same process, except much longer arcs.)

[Gary_S]

> touch the black probe to

No. Well, that works, but---

As HBP said (and you quoted): If you try to do any part of this without clipping the leads in place, you WILL get arcing/sparking that will startle you, make noise, and let you know you're all primed for electrocution.

When you 'touch', the tube current in the winding diverts to the "short cut" through the meter. When you un-touch, that current diverts back to the winding.

When you instantly start/stop current in a large winding, there is a huge voltage spike. As you pull the meter probe off, the voltage kicks, arcs back to the probe, arc breaks, kick again, re-arc. (If you've used an old-fashioned welder, same process, except much longer arcs.)

You'd have to use the clip on test leads with an alligator clip instead of the probe?

Does anyone on the site use this method? It would be interesting to hear their views on this if anyone does use this as their main biasing method.

[darryl]

Does anyone on the site use this method?

I certainly don't. The 1Ω cathode resistors are simple, cheap, permanent and safe.

You have now received the same message from six el six, PRR, tubenit, HotBluePlates, alerich, sluckey and darryl.

Why persist with this enquiry?   

[tubenit]

PRR, HBP & Sluckey are the smartest amp guys I know on any amp forum. They have conveyed this is a dangerous method.

I have been on the forum for yrs and yrs, and I don't know anyone who uses this method nor remember anyone personally advocating it.

In my opinion, it is a dangerous, unnecessary and foolish method of measuring bias. You have been given more than adequate information and unanimous warnings from every responder to your inquiry.

With respect, Tubenit

[phsyconoodler]

There is another easy and relatively safe way of measuring bias.
  Take your meter in an unplugged amp and measure from pin 3 of octal tubes to the center tap and record the ohms reading for each tube or grouped pair.Then turn the amp on and measure between the same two points,only measure DC voltage this time.All done with alligator insulated clips of course.
  Then simply divide the ohms into the voltage to get the current in milliamps.
e.g. say a Fender amp has 49 ohms on one leg and 52 ohms on the other leg( not uncommon to have a slight imbalance) 1.5v/49ohms=.0306 or 30.6ma.now the other side: 1.5v/52 ohms= .0288 or 28.8ma.

Easy and safe as long as you use insulated and very high-quality clips that won't slip off.
The voltage you read here is very low but you still have to be careful.

[cbass]

Thats the way I do it.I'm not smart enough to use insulated clips though.I would recomend everyone else does though
Its the easiest method for when I don't want add 1 ohms(vintage amps)

[Gary_S]

Thanks psychonoodler and cbass for the info, i appreciate you replying.

Cheers guys.

[PRR]

I had in mind the Hewlett-Packard 428A. That's 1958 hi-tech,...

[Dave]

Let me try to put this in simpler terms. If you set your meter to measure DC current and you touch one probe to the OT center tap with the amp running, the other probe becomes a welder and if you touch it to your skin, it becomes the Grim Reaper.


Dave

[dude]

I’ve done the shunt method once, just like phsyconoodler mentioned, it works.  l used insulated clips but once was enough and the amp was a Blues Junior, had every connection far apart and easy to get at. The 1 ohm R from pin 8 to ground gave me just about the same bias, l usually subtract about 2mV for the screens. For what it’s worth, l can’t hear any difference in 2mV and my tubes plates probably can feel any difference either  .


I like HBP’s mention of using a 10 ohm R rather then 1, 1 ohm R are at best 2%, if you can find them, while 10 ohms can be much more accurate, just use the first two numbers for mV’s.
In an amp like you’re describing, l’d totally pass on the shunt method, follow the advice given, might just save your life.

[AlNewman]

There should be no reason to set your multimeter to DC current.  Everything can be done with DC volts to chassis, with a measurement of the output transformer resistance.

This is with guitar amplifiers and output transformers.  I have no experience with a Hewlett Packard.

[PRR]

GUYS, you last three posters: This thread died eleven years ago. I am sorry if I confused you.

[mresistor]

Well dang it PRR!   

[WimWalther]

In any case..

The shunt method does work, but it introduces two types of error:

1) It only accounts for plate current (Ia), screen current is not taken into account.

2) The OT creates a parallel (to the DMM) current path which causes the Ia reading to be slightly lower than actual.

But for quick checks of bias current, it does give a generally good idea of how things are set.

As has been pointed out by PRR and others, the best way of measuring bias is as cathode current (Ik) via a low-value current viewing resistor in the cathode circuit. 1R 1% 2W parts are ideal for this.

[sluckey]

The shunt method does work, but it introduces two types of error:

1) It only accounts for plate current (Ia), screen current is not taken into account.

That's not an error! When you want to calculate plate dissipation, you don't want screen current to be a part of the current. Using the cathode resistor DOES introduce an error because the screen current is part of the cathode current.

[WimWalther]

The shunt method does work, but it introduces two types of error:

1) It only accounts for plate current (Ia), screen current is not taken into account.

That's not an error! When you want to calculate plate dissipation, you don't want screen current to be a part of the current. Using the cathode resistor DOES introduce an error because the screen current is part of the cathode current.

If you're only wanting to measure Ia, then I can't really disagree!

[sluckey]

If you're only wanting to measure Ia, then I can't really disagree!

Yes, that's the whole point. All you want is Ia when you want to calculate plate dissipation in order to set the bias voltage to give some magic number like maybe 70% max PD. Using the cathode resistor does introduce an error but I gladly accept that error just because this method is safe and easy. But keep in mind that screen current in some tubes such as EL34s can be significant. If I owned any EL34 amps I would still use the cathode resistor method but I would also measure the voltage drop across the screen resistor in order to calculate screen current and subtract that from the cathode current.

[shooter]

^^^ agreed, i "accept" the error as erroring on the safe side setting bias.  If I set bias based on both G2i and Ia = plate dissipation, then my bias will be cooler, allowing for the variance in guitar hammering time over all musicians.


and, yes EL34's are an exception to the rule but they seem to be in amps used specifically to be abused beyond normal operating range 

[HotBluePlates]

These days, I go for the Eurotubes Pro One.
It measure plate current only, plus plate-to-cathode voltage.  And if you pay a little extra, it calculates plate dissipation for you.

I don't need it to get the job done, but it's a lot easier if the money is available to spend.

[wsscott]

Would you describe the process you go through when using the 1 ohm bias method? How do you hook up the DMM, then what do you do with the readings? 

And how do you bias if you have a simple bias probe that measures only the plate current?

[sluckey]

read...

     https://www.duncanamps.com/technical/lvbias.html

[wsscott]

Thanks.

[WimWalther]

Would you describe the process you go through when using the 1 ohm bias method? How do you hook up the DMM, then what do you do with the readings? 

If you have a 1R resistor in the cathode circuit, you simply read DCmV across it and the reading converts directly to mA.

70mVDC = 70mADC. This gives total current through the tube, so the sum of both plate & screen.

And how do you bias if you have a simple bias probe that measures only the plate current?

If your bias probe reads only plate current, then set it per the datasheet value, usually between 75% and 90% of max for an AB amp.

If individual tubes have screen resistors, measure screen current as voltage drop just as with the cathode resistor, but with a conversion rate based on the screen resistor value.

[dwinstonwood]

...usually between 75% and 90% of max for an AB amp.

Some might say that's pushing it for AB1 guitar amps (not so with Class A). Especially when what we all really want to hear is all knobs at 10. 

[glass54]

My preference is 1R0 1%. Measurements will always be within 10% which equates roughly to Screen current.
....and this is a safe, easy to read method with little risk of accidental fireworks 
As an option which I might occasional resort to, the lowest maximum current rated True RMS clamp meter (which I own and recommend):
https://au.element14.com/tenma/72-2985/dmm-clamp-auto-100a-3-75-digit/dp/2667180
No removing wires, very safe, etc. You can also measure your AC currents in Power Transformer Primary cct, etc, etc.
My thoughts on the matter.
Regards
Mirek

[WimWalther]

...usually between 75% and 90% of max for an AB amp.

Some might say that's pushing it for AB1 guitar amps (not so with Class A). Especially when what we all really want to hear is all knobs at 10. 

Opinions are certainly welcome. I don't see 75% of max being that awfully high.. 90%, maybe so. I suppose its a matter of how deeply into the A phase of the AB range we wish to go.

A lot of commercial AB hifi amps are biased very shallow, as in you're lucky to get even 10% of full output from the A range. It all comes down to how much heat you're willing to deal with - both in the tubes and the PT.

Of course, most amps tend to sound their best when they are run right to the ragged edge of self-destruction.. so.. =)

[wsscott]

Good thought on the clamp meter.  I have one that I could use.  Just never thought of using it for this stuff, but seems like a good idea.

[pdf64]

... set it per the datasheet value, usually between 75% and 90% of max for an AB amp.
 ...

Bear in mind the rating system applicable to the info being referred to, and how the different rating systems are intended to be used.
eg 6V6 anode limit is 12W design centre, 14W design max.

Design centre limits apply with relevant stuff such as mains voltage and load impedance at their nominal level.
Design max limits apply with such variables all at their worst likely level, eg mains voltage at its upper allowable level, load impedance possibly halved if an extension speaker ouput has been provided with no provision for changing the OT ratio.
90% of 14W is >12W.