Stand by + series R's

[rzenc]

Howdy!

I'm currently building a 100W and would like to hear what you guys think about the idea I drew (see attached schema)

As I see it: at initial turn on, reservoir and subsequent PS caps will be short circuits and with series R's added (labeled R01 and R02), it will limit how much current can flow.
H.T. winding has 370 - 0 (Bridge recto type) which unloaded might gimme 518VDC. Reservoir cap is two 100uF series + balancing resistors.

I estimated 15Kohms/50W for each part of winding added series resistance. It could limit current inrush for about 34mA's.

Now things go tricky... when in stand by mode, 15K will be in series with B+ rail and is supposed to drag it down...but how much voltage will be dropped? I still did not figured it out how to adjust series R's since tubes will be dragging current from reservoir and PS caps as well....

Can you guys help me??

Thanks in advance!

Best Regards

R.

[sluckey]

What is your goal? Limiting inrush cap charge current? But why? To protect the PT, or caps, or something else?

It's true that at time zero (T₀) cap charging current will be at some max value, limited only by the series resistance involved. But T₀ is just a brief point on a time line. Even at 1µs past T₀ the cap voltage is rising and the current is falling. Caps charge at an exponential rate (natural log function, LOG_e). Charge voltage starts at zero (current starts at some max value) at T₀ and increases at a Log_e rate and will be fully charged after 5 RC time constants. The charge rate is very rapid initially but more slowly as time progresses. Looks just like a natural log plot.

I'm not sure anything in an amp would benefit from such an arrangement. But I think that what you want would not be a simple single throw switch, but rather, a switch that has 3 positions, ie, full standby, intermediate current limit, and full operate. A 3PDT would do it.

[stratele52]

Can't read your schematic ; file damaged  ?

[kagliostro]

I can read the schematic, no problem

About the idea, if you want a standby and so you insert a switch, I think ......... why not ?

Also Sluckey version is to be tried

K

[drgonzonm]

It looks like you are using a stereo jack to power your xfrmer.  Are you feeding the transformer with two hot legs? 

I am concerned that the jack will not stand up to amperage that the fusing is suggesting. 

Then the fusing scheme on the secondaries seems a light,  for example, fusing 5 amp on the primaries, so shouldn't the fusing on the secondaries about 5/3 or 1.5 to 1.75 amps.  (You may want to look at 460 to 480 volt fusing, unless you have a source for 400 v fuses.) 

How are you planning to power your valve heaters?

[rzenc]

What is your goal?

Keep some voltage on tube plate.

This amp is going for a recording studio and I know the guys for a long time...Did some sessions together...They will probably leave on stand by for days. 

I read some info about cathode poisoning if tubes are left warm without current flow. Immediately this amp build came to my mind. The studio owners are very cool guys and willing to use exotic tubes. I will feel better knowing I did whatever possible to extend tube lifespan.


So, here's the main idea: when in stand by mode, it would have around 200V @ power tube plates and less on following stages.

Say, underload, Voltage @ power tube plates is 480V. EL34 based amp. 1750K primary Z. Plate Diss @ 25W = 52mA's. 4*0.052=0.208A=208mA's.
When engaging series R's, this current should flow, but as voltage goes down, plate current will also drop and since bias will remain the same negative voltage it will also force plate current down, thus keeping it at some lower setting, but still conducting some current and hopefully will help to extend tubes life.

The question is how to determine the value of the resistor and it's power dissipation.

It looks like you are using a stereo jack to power your xfrmer.  Are you feeding the transformer with two hot legs? 

I am concerned that the jack will not stand up to amperage that the fusing is suggesting. 
How are you planning to power your valve heaters?

No. I'm not using stereo connectors for it. Drawing Program names it wall outlet connector.

I drew only H.T. Winding. The rest of the windings would only mess with it.


Regarding heaters, power tubes will be 6.3VAC and preamp 12VDC.

Thanks for the insights so far!!
Best Regards!

R.

[HotBluePlates]

According to "Getting the Most from Vacuum Tubes," the issue described was called "interface resistance" in the old days. The way "cathode poisoning" is sometimes described these days, it is more a misapplication of a problem that occurs in transmitting tubes and not the types or voltages we use in guitar amps.

The problem was caused by no tube current during standby, so lowering plate voltage and/or limiting tube current seems to be a poor solution.

One option is to reduce the heater voltage and therefore power during standby; but this might require some circuitry complications in the heater supply.

A different option is to ensure the tubes always draw some amount of current. This will happen anyway if the amp is left on with no standby or other trickery. Note that this suggestion came from Tomer, who wrote the book I'm citing, and who was an engineer for a major tube manufacturer in 1960 when the book was written. Tomer wrote that interface resistance typically occurred in tubes specialized circuits such as blocking oscillators, which are not found in guitar amps, and where the tube may be far into cutoff for very long periods before passing large current pulses.

So the best option seems to be to leave the amp on, heaters at regular voltage/current and B+ at normal levels to ensure some idle current draw at all times. If a standby/mute function is needed, it might be best to use a method to mute the audio output, such as shorting the grids of the push and pull sides of the output section together.

You might ask whether this makes sense, or if the tubes will wear out faster. I, and others, have had tube audio power amps and test equipment that was powered uninterrupted, for probably years. I've noticed some of the pieces I've bought have the original tubes installed by the manufacturer (known because they were prominent tube makers, but with the equipment company's name stenciled on the tube). Almost universally, when I've pulled these tubes and put them on a Hickok tester, they've all tested as good tubes with plenty of life left.

So your choice: convince your friends to shut things down daily, or just insure the tubes always draw some current.

[PRR]

> "interface resistance" in the old days. The way "cathode poisoning" is sometimes described these days, it is more a misapplication of a problem that occurs in transmitting tubes

I thought it was first annoying in _computer_ tubes?

If a bit holds a "zero" for a long time, B+ is on the tube but the grid is all the way to cut-off.

When a "one" finally happens, grid comes up near cathode, not-cut-off, *sometimes* the tube did not react strongly right away.

And most of these tube computers were *very* marginal things. Tubes are more expensive than under-paid grad-students (the main alternative computer). All kinds of mickey-mouse logic tricks were used to reduce the number of tubes, logic swings were weak.

So if a long-off tube were weakly ordered to tune on, sometimes it didn't.

I think this was traced to certain unnecessary elements in cathode Nickel and in the oxides. New specs made interface largely vanish. The Nickel specs would have applied to "all" tubes; the tube industry's consumption of Nickel was such a small part of overall Nickel production that the Nickel industry didn't want to fool with multiple specifications, the tube producers had a committee agree on a spec they all could use.

Not sure what Nickel the Russians and Chinese use now that computer-use is obsolete. But you say NOS will be used.
______________________

Back to the real question: I think 200V even 100V on plates is hardly "standby", the tubes will play a lot louder than a radio even though not to full 100W Marshall expectation.

> how much voltage will be dropped?

Depends on the tubes (mostly the big ones) and their bias.

If self-bias, the tubes can be approximated as a resistor. Assume typical self-bias two-6V6 running 300V 90mA. 300V/90mA= 3.3K Ohms. If you want to drop 300V to 100V, put 6.6K in series. You have 300V-100V= 200V across 6,600 ohms, 6 Watts.

This assumes drop-resistor after the first power cap. Resistors before the first power cap see a very strange waveform, calculation is a pain. For a studio amp, you probably may as well ought to take the power tubes from a C-R-C filter for lowest-hum/buzz. The R could be a hundred or so ohms for full power, 5K-10K for low-low-low power.

In "fixed" bias, if G1 voltage is held nominal and G2 voltage drops, plate current drops a lot. This is the same condition which once caused "interface" trouble in computer tubes. While I doubt it is a big problem, I really don't see the point.

Leave em on for a few days. Or kill the B+ and leave the heaters on. Or kill the whole thing. I don't think it is a big difference.

If you really loved the tubes you'd keep B+ well below 514V. A "recording studio" does NOT need actual 100 Watt POWER, just "that hunert-watt Marshall sound". If it's 108dB instead of 112dB in the room, well, the engineer has a knob for that.

[smackoj]

"If it's 108dB instead of 112dB in the room, well, the engineer has a knob for that."

MIT is looking for an engineering grad program director with a sense of humor.....i nominate P-R-R

 

[rzenc]

This assumes drop-resistor after the first power cap. Resistors before the first power cap see a very strange waveform, calculation is a pain. For a studio amp, you probably may as well ought to take the power tubes from a C-R-C filter for lowest-hum/buzz. The R could be a hundred or so ohms for full power, 5K-10K for low-low-low power.

In "fixed" bias, if G1 voltage is held nominal and G2 voltage drops, plate current drops a lot. This is the same condition which once caused "interface" trouble in computer tubes. While I doubt it is a big problem, I really don't see the point.

Ok, got the idea... let's give a try and I will report back...

For a studio amp, you probably may as well ought to take the power tubes from a C-R-C filter for lowest-hum/buzz.

I could lower background noise on latest preamp build by running heaters @ 12VDC and careful layout. It has C-R-C-R-C to C.F. which made it very quiet indeed.

A "recording studio" does NOT need actual 100 Watt POWER, just "that hunert-watt Marshall sound". If it's 108dB instead of 112dB in the room, well, the engineer has a knob for that.

Let me kindly disagree, but there is no way a 15W or 30W to sound like a 100W beast. Pair it with some amazing speaker cabinets and a Neumann TLM 170R and you have thick metal tones. These guys are a lot into recording metal and other extreme stuff. The way such a monster pumps low end is massive. And as you know, it sounds absolutely different from turning up mixing console gain or bass pots...

Thanks guys!

I will report back results.

Best Regards

R.

[HotBluePlates]

> "interface resistance" in the old days. The way "cathode poisoning" is sometimes described these days, it is more a misapplication of a problem that occurs in transmitting tubes

I thought it was first annoying in _computer_ tubes?

Yeah, I kind of mixed-metaphors, between interface resistance in computer tubes and cathode poisoning in transmitter tubes, which then necessitated a standby.

Ultimately, what I was getting at is Rzenc's situation doesn't fit either of these categories, and the whole thing is really a non-issue (though some writers make it seem otherwise).

A "recording studio" does NOT need actual 100 Watt POWER, just "that hunert-watt Marshall sound". If it's 108dB instead of 112dB in the room, well, the engineer has a knob for that.

Let me kindly disagree, but there is no way a 15W or 30W to sound like a 100W beast. ... The way such a monster pumps low end is massive. And as you know, it sounds absolutely different from turning up mixing console gain or bass pots...

He really means that as long as the "talent" doesn't get upset about it, it makes no difference if it was a 50w or 100w amp.

O'Connor pointed out there was some (maybe unintentional) trickery related the feedback loop that gave 100w Marshall's their character. He also provided a project in one of his books that outputs 50w of usable power, but uses 4 output tubes and makes use of the features of the 100w amps. I never tried it, as I have no need for an amp even 1/3 as loud as that. I can also think of better ways (for me) to use 4x output tubes.

That said, I once recorded a band that cranked a 50w Marshall. I got a huge sound for the solo, because even a 50w amp was overpowering the iso-booth of a multi-million dollar studio, and bleeding through "soundproof" doors at shockingly high volume. I put up a close mic on the cab and a 2nd mic 30-40ft down the hall from the booth's door.

It sounded great, but I probably could have done just as well with less volume from the amp and just use the enormous EMT plate at the studio.

[rzenc]

This assumes drop-resistor after the first power cap. Resistors before the first power cap see a very strange waveform, calculation is a pain. For a studio amp, you probably may as well ought to take the power tubes from a C-R-C filter for lowest-hum/buzz. The R could be a hundred or so ohms for full power, 5K-10K for low-low-low power.

You got me thinking about it..   

50uF + 250R sets low roll of @ 12.7Hz. When steady, power tubes might be biased @ 50mA each...so 200mA's would flow thru dropping R + a few mA's...:
250R*0.2=50V dropped on R. 50V*0.2A=10W dissipate on R.

However, How many dB's would it reduce buzz/hum @ power plates?

I can also think of better ways (for me) to use 4x output tubes.

Come on HBP...we are looking forward to your idea!

Thanks guys!

Best Regards

R.

[PRR]

You already have the numbers.

> 50uF + 250R sets low roll of @ 12.7Hz.
> How many dB's would it reduce buzz/hum @ power plates?

The buzz is 120Hz.

The R-C product is 12.7Hz.

120Hz is reduced 12.7/120 or to 0.106 of before.

I know 0.1 of voltage is -20dB. I'll let you figure 0.106 if you care. (However the 6% difference is lost in 10% resistor and 20% cap slop.)

[HotBluePlates]

What I meant was O'Connor has a design where an amp makes 50w of output power, but the output stage uses 4x EL34's.

It's not class A, the extra tubes are used to effective boost the Gm of the output stage and give the feeling of a bigger 100w amp without the extra volume. I never bothered to build it, so I can't report on whether it's useful.

I did own at one time a '73 Marshall 50w head and an early-70's 100w stright cabinet, as well as a Hiwatt DR504. A punk band used my Hiwatt and the cabinet to record an album.

Anyway, I can't imagine having a 100w version of either of those amps cranked in a studio. Ears would bleed and I'd worry about blowing out a condensor mic. Bad things happen when the capsule skin bottoms out against the backplate.

[rzenc]

I know 0.1 of voltage is -20dB. I'll let you figure 0.106 if you care. (However the 6% difference is lost in 10% resistor and 20% cap slop.)

Found this: Gain(dB)=20*log(Vout/Vin)??

0.106 is ratio between cut-off frequency and ripple frequency, times 20log = -19,5dB ??

I'd worry about blowing out a condensor mic.

HBP, check it out: http://www.neumann.com/?lang=en&id=current_microphones&cid=tlm170_data

This is the mic I was talking about early. It can withstand about 154dB SPL @ 1m. And it's such an amazing secret weapon when in need to record High SPL's. To close the deal, finding a preamp to match what sound you are looking for. Tried and true many times with insane amps crancked...althought by crancked I mean around 1 to 2 pm ( marshall would label it as 7 - 8 on the pots ), not 10...

Give this mic a try next time you need to record High SPL's while trying to capture what's really going on inside the booth, not trying to build a sound different from what is being pumped out from the speakers.

Thanks guys!!!

Best Regards

R.