Too Much B+ voltage

[Lectroid]

I've been building a 6BM8 design suggested by tubenit. I wanted a quiet push-pull Marshall-ish amp around 6-7 watts, with reverb and a little overdrive if possible. His design and my version are attached.  I followed it mainly except I left out his MOSFET cathode follower.

The problem is that the B+ voltage, and all my anode voltages are much higher than on tubenit's schematic. Here's what I measured:
   A: 384   
   B: 378
   C: 313
   D: 302

Anode voltages:
   1st triode: 197V
   2nd triode: 199V
   PI triodes: 240V and 204V

The PT is from a Hammond A0-43 chassis, with 325-0-325V, and a 5AR4/GZ34 rectifier.

At that point I stopped testing before I broke something, and came here to get some help.

Are my B+ values too high for this 6BM8 design?  Or--will the high B+ rail values drop down far enough, once the tubes are under load? Am I okay going ahead with injecting a signal and continuing with startup? 

Should I give up and get a new PT? Or would it help to replace the 15H choke with a dropping resistor, or to replace the 5AR4 with a 5Y3?  Would the 5Y3 produce enough current?

Thanks in advance,

[tubeswell]

B+ voltage,... measured:
   A: 384   
   B: 378
   C: 313
   D: 302

That should drop when you pop the 6BM8s in (Cathode bias always knocks heaps of B+ off)

[Lectroid]

That should drop when you pop the 6BM8s in (Cathode bias always knocks heaps of B+ off)

.

@tubeswell,

I should have said: "All tubes were inserted when I made those measurements."  Including the 6BM8's.

Sorry it wasn't more clear in my post.

[sluckey]

Hmm, tubenit used a PT with 190-0-190 winding. You're using a PT with 325-0-325 winding. What's wrong here?   

[Dave]

You say you are using a 5AR4 but your schematic says 5Y3.


Big difference there on B+ voltage.


Dave

[Lectroid]

Hmm, tubenit used a PT with 190-0-190 winding. You're using a PT with 325-0-325 winding. What's wrong here?   

@sluckey,
Sure, I knew that the PT might be too big for the amp, but it was worth a try since I already had the transformer.  And--it doesn't work.  In this case you think I should just get the smaller PT.  Thanks. 

But are there practical limits to substituting one PT for another in a given circuit?  Is a 10% B+ over-voltage allowable?  20%?  Just curious.

[tubeswell]

Okay - so what's the cathode current (measured across the cathode resistor) - You'll need to ohm-out the exact resistance of the resistor and divide the resistance from the voltage drop across the resistor to get the current. (If the 6BM8s are still way under biased - i.e. if they're not already at 100%, you could decrease the cathode resistor size and get them running at 100%). (You haven't given enough info about the set up yet). (Alternatively, if the tubes are running too hot, you could increase the cathode resistor size and cool them down a bit


Edit: V/R = I

[PRR]

> Is a 10% B+ over-voltage allowable?  20%?

But you tried a 70% over-voltage.

That's a lot like driving a 13-foot-tall truck under an 8 foot high bridge.

Yes, 190VAC is a VERY odd or uncommon transformer voltage, but you'll probably regret anything very different.

Tubeswell is starting to tell you something but I don't follow R*V. IAC blowing-off 70% of your power is poor form.

[tubeswell]

A 5Y3 rectifier might also knock some B+ off (or not)

[Lectroid]

Thanks everyone,

I understand that this PT isn't suitable, that it's way overpowered.  Got it.  I'll make appropriate changes.

I'm still hoping someone might explain how to know whether a potential replacement PT is  suitable.

Or an answer to my other question about how the same PT can be used in such a range of Fender amps.  That one baffles me.

@tubeswell,
When I first fired it up, I calculated the bias per Robrob's cathode resistor method and got 125%.  I raised the cathode resistor up to 880 ohms to get to 60% bias.  That brought the bias voltage to -32V.

[tubeswell]

@tubeswell,
When I first fired it up, I calculated the bias per Robrob's cathode resistor method and got 125%.

What was the B+ voltage at that time?

I raised the cathode resistor up to 880 ohms to get to 60% bias.  That brought the bias voltage to -32V.

What did the B+ voltage go to after that?

[HotBluePlates]

I'm still hoping someone might explain how to know whether a potential replacement PT is  suitable.

The PT delivers AC voltage, but we will convert to DC voltage.  The filter caps tend to charge to the peak of the PT's AC voltage waveform.

The PT's AC voltage is described in "Volts RMS" of a sine wave, and we assume the outlet voltage is a sine wave.

The linked article goes to some lengths to figure out what "RMS Volts" is of a wave whose instantaneous voltage we know at a number of points in time.  It derives a formula for calculating the RMS voltage of the sine wave when we know the peak voltage of that wave, but we can go the other way:  RMS Volts x √2 = Peak Volts

When we rectify the PT Voltage, we will get a DC Voltage equal to the PT Voltage's peak volts, minus some amout due to losses across the rectifier, or because the amp circuit is sucking current out of the filter caps.  However, it's a start to know whether a transformer is suitable.

Tubenit's PT had a PT delivering 190-0-190v, so the peak voltage output is 190v x √2 = ~269v DC.  This is exactly the value he indicates at Point A of his schematic.  Tubenit probably got more like 268v, because there's about 1v of loss across the solid-state rectifier diodes he used, but this discrepancy is immaterial.  A tube rectifier will have a much greater voltage-loss, but the calculation is very involved such that many folks just use rough-guess, or use various fudge-factors to estimate the reduction from peak voltage.

To estimate a suitable PT voltage when the desired power supply DC voltage is known, you can use the article's final formula to calculate RMS Volts from a known Peak Voltage:  Peak Volts x 1/√2  =  Peak Volts / √2  =  Peak Volts x 0.7071.

For Tubenit's circuit, we calculate 269vdc x 0.7071 = 190.2v RMS  ----> we use a bridge rectifier with a single winding of 190v AC, or we use a full-wave (not-bridge) rectifier of 190-0-190v AC.

[tubenit]

I'm still hoping someone might explain how to know whether a potential replacement PT is suitable.

I use this tube data cheat sheet attached.

As an illustration ........................ let's say I wanted no more than and around 300 volts on the plates of 6BM8's.

If I was using a solid state rectifier then I'd divide 300 by 1.4 and get 214.  So, I'd probably look for a 210-0-210 PT

If I was using a 5Y3GT, then I'd divide 300 by 1.1 and get 273.   So I'd probably look for a 270-0-270 PT

If I were using a GZ34, then I'd divide 300 by 1.3 and get 231.  So, I'd probably look for a 225-0-225 PT

And of course you'd make sure the PT rating for milliamps is adequate/proper for the tubes both on plates and heater wiring.

Anyhow, that's my starting place to think thru PT selections. So far, this has served me well.

With respect, Tubenit

[tubenit]

So, I've built two amps in the last few yrs using 6BM8's.  I wanted plate voltages under 305v based on a previous build with 6BM8's.


I used solid state rectification on both recent 6BM8 builds to minimize chassis space. (hence "the Minimalist")

One amp was a 210-0-210 PT   (x1.4) suggests a plate voltage of 294v.  In real life, the volts were 284 not subtracting the cathode voltage from that.

The other amp had 190-0-190  (x1.4) suggests a plate voltage of 266v.  In real life, the volts are 269 last time I checked.

This tube cheat sheet approach is just a starting place to begin looking for what you need in the way of a PT both in voltages and the ma needed.

with respect, Tubenit

[Lectroid]

tubenit, HotBluePlates,

Thanks for explaining the process of evaluating a power transformer for me.  I understand how to approach it now.

Now, if only someone could tell me how the Champ and the Princeton Reverb can both work using the same PT, I'd be a happy solder jerk.  One is a simple two-tube amp, the other is a six-tube complex amp with effects.  How can they both be happy running the same PT?

I love learning this stuff, thanks again!

[pdf64]

…
Now, if only someone could tell me how the Champ and the Princeton Reverb can both work using the same PT, I'd be a happy solder jerk.  One is a simple two-tube amp, the other is a six-tube complex amp with effects.  How can they both be happy running the same PT? …

What will do a lot will do a little.
Your car is happy to go, whether there’s only the driver, or 4 passengers and a full boot.

[Lectroid]

What will do a lot will do a little.
Your car is happy to go, whether there’s only the driver, or 4 passengers and a full boot.

@pdf,
Are you equating a car's available horsepower [motive force], to the total B+ available in an amp?  It sounds like you're saying that "the car has X available horsepower but will only use what it needs depending on the load it has to carry?"  Have I got that right?

That seems to imply that the Champ only "pulls" the B+ voltage that it wants from the 125P1B PT, while the PR just "pulls" more.  That doesn't seem to align with what everyone else said above.   

I resolved my confusion about how those Fender amps worked.  They used different combinations of rectifiers, and the primary winding resistance of different OTs, to get different B+ values while still using the same PT.  Elegant and probably made the bean counters happy, too.

The Champ used a 5Y3 to get a lower plate voltage (360V) with the 125P1B PT; on the PR (420V) they used the 125P1B with a 5AR4 and a different push-pull OT.  The Bronco used the same 5Y3  but then ran the B+ through a different single-ended OT which dropped the voltage to 342V on the 6V6 plates.

Thanks to everyone who chimed in here to fill ignorance with good knowledge.  You're the best.   

[Lectroid]

I'm still hoping someone might explain how to know whether a potential replacement PT is suitable.

I use this tube data cheat sheet attached...

@tubenit,
Thanks a lot for this.  I have the cheat sheet already but your post gave me real practical tips on how to approach the design problem of selecting/substituting a PT.  It was exactly what I was asking for.  Thanks again for your help.

If I could ask a question, how did you select the plate voltage of 269V for the 9W 6BM8 amp?  Or was it a case of having the PT, figuring out what the likely B+ was going to be, and finding a tube that would work well in that voltage range?  And what was the bias voltage?

@HotBluePlates,
Thanks to you, too, for laying out the basic math needed to puzzle out a new design with a new PT or a substitute.  Between you and tubenit, I now have a good set of tools to think about this.  Thanks again.

[Lectroid]

@tubeswell,
When I first fired it up, I calculated the bias per Robrob's cathode resistor method and got 125.

What was the B+ voltage at that time?

I raised the cathode resistor up to 880 ohms to get to 60% bias.  That brought the bias voltage to -32V.

What did the B+ voltage go to after that?

I can't be sure, still learning how to do startup.  The final plate voltage was 368V (418V from the rectifier) with the -32V bias.  (I know, should be closer to -16V). The rectified DC was 418. 

The cathode current after the resistor replacement was .036A. (32V/880 Ohm). Robrob calculates the plate dissipation at 62% of max.

I made my own mistakes in this build, no reflection on tubenit.

Thanks, hth

[ac427v]

I am a seat-of-the -pants builder so take this with a grain of salt.

Before ordering a new transformer you could try one of the ideas you asked about in your original post.
Replace that 14 h choke with a dropping resistor and measure all the voltages and plate dissipation again. If it is close to your goals, try playing through it.
I would use a 2.2k or 4.7k metal oxide resistor. Rated for at least 2 watts. You might decide it sounds too "tweedy". Or you might like it a lot.

[mresistor]

You can also use a 50 watt zener on the PT HV center tap to ground to drop a lot of volts depending on the volt rating of the zener.  They're not cheap though.


https://www.ebay.com/itm/383376555783?chn=ps&norover=1&mkevt=1&mkrid=711-213727-13078-0&mkcid=2&itemid=383376555783&targetid=4580496737259386&device=c&mktype=&googleloc=&poi=&campaignid=418233788&mkgroupid=1230353745471221&rlsatarget=pla-4580496737259386&abcId=9300542&merchantid=51291&msclkid=2aef27ee02f11596888542914ee9781b

[PRR]

...tell me how the Champ and the Princeton Reverb can both work using the same PT, I'd be a happy solder jerk.  One is a simple two-tube amp, the other is a six-tube complex amp with effects.  How can they both be happy running the same PT?

(Say) I have an 8-room house, and neighbor George has 4 rooms. Two people here, one person there. We like light. We can very well use about the same street transformer and fuse box.

They both eat the same/similar voltage.

The Prince sucks twice the current (both B+ and heater) because two big tubes. (Like night-lights, little tubes rarely add-up to much.) So the Champ "can" use a smaller PT.

The smaller PT "should" cost less. But the labor hardly changes, so the price is not a lot different. Especially when you order them by the truck-load.

In mass production it is distressing to run out of a part. If one part serves two models, you buy for both and are less likely to run out. Especially if total sales of Champs and Princes is about constant but some years you move more of one, more of the other. (Guitar-god appearance on BandStand or Popular Guitar magazine..., or economic trends where another 10 bucks is/is-not important this year.) So at some point Leo the bookkeeper realized that buying all the same part was the best business. (This thought may have come to 1930s radio makers who had a basic 1-6V6 and a fancy 2-6V6 radio and coud not always predict their market balance.) While this may be wrong in one-off or small production, today nobody seems to care, they sell the 2-6V6 for all small Fender-clones.

[Lectroid]

@ac427v,
Thanks, I always like simple solutions.  I may do that just for fun, since they aren't a lot of 190-0-190V PTs for sale. I did get the impression from some of the others that there was no way to make that PT work w.er-voltage.  But it wouldn't hurt to try.

@mresistor,
Thanks. Using a Zener to drop 50 inconvenient volts is too much for my purist sensibilities.      Anyway, I need to lose about 100V.  I might just adapt this into a 6V6 Plexi.

[PRR]

https://primarywindings.com/product/amp-maker-pwam03d-190v-0-190v-80ma-6-3v-2a/
(very English, £28-£48 shipping to USA)

https://www.aliexpress.us/item/2255800144728597.html
(Note shipping!)

[Dave]

Now, if only someone could tell me how the Champ and the Princeton Reverb can both work using the same PT, I'd be a happy solder jerk.  One is a simple two-tube amp, the other is a six-tube complex amp with effects.  How can they both be happy running the same PT?

Also, the voltage is lower on the Champ because of the 5Y3 rectifier.


Dave

[mresistor]

@ac427v,
Thanks, I always like simple solutions.  I may do that just for fun, since they aren't a lot of 190-0-190V PTs for sale. I did get the impression from some of the others that there was no way to make that PT work w.er-voltage.  But it wouldn't hurt to try.

@mresistor,
Thanks. Using a Zener to drop 50 inconvenient volts is too much for my purist sensibilities.      Anyway, I need to lose about 100V.  I might just adapt this into a 6V6 Plexi.

there are 100v 50w zeners

[tubeswell]

The final plate voltage was 368V with the -32V bias.  (I know, should be closer to -16V).

The cathode current after the resistor replacement was .036A. (32V/880 Ohm). Robrob calculates the plate dissipation at 62% of max.

You can easily decrease that cathode resistance some more and it should bring the plate voltage down a bit. For these amps in cathode bias, don't be shy to run the tubes at 100% of Pmax - they will be fine.

[Lectroid]

@PRR,
Thanks for your comments on the Champ/Princeton question.  I'd already answered my own question by comparing the amps' schematics, in reply #16.  They did it by using various combinations of rectifier tubes and OT winding resistance to get the various B+ levels out of one PT model.  But it's good to get validation about my musings on the Fender parts supply model. 

Also thanks for the links to the 190-0-190 transformers.  That was very helpful

A question: how to know if a PT can supply enough current for a given tube? The English PT puts out 80mA.  Is this enough for this circuit?  For the 6BM8, Max Plate Current plus Max Screen Current = 42mA.

For these amps in cathode bias, don't be shy to run the tubes at 100% of Pmax - they will be fine.

@tubeswell,
I'll keep that in mind.  Would that push the tube's max power dissipation?  Would it shorten tube life at all?

[tubeswell]

For these amps in cathode bias, don't be shy to run the tubes at 100% of Pmax - they will be fine.

@tubeswell,
I'll keep that in mind.  Would that push the tube's max power dissipation?  Would it shorten tube life at all?

Pmax is Pmax. In the old days, the manufacturer's data was contrived so that tube operators would run tubes so that they wouldn't give out before the warranty period, or you would void the warranty. So theoretically, you ought to be able to run the tube at Pmax and it will last for the recommended number of hours on the tube data sheet before it needs replacing (but its all about 'averages'). For the most part, modern manufacturers just 'copy' old tube designs and old tube data sheets, so who knows?. On the face of it, the tube is designed to safely run at Pmax, (and modern tube manufacturers are cashing in on old tech). Besides that, tubes are inherently more rugged electronically than transistors, so they can take a fair bit if 'abuse'. Having said that, running a tube at Pamx is not really 'abuse' - rather, its running the tube within the stated design tolerance. For cathode-bias, the cathode resistor 'auto-biases' the tube anyway - if more current flows through the cathode, Ohms Law is at work through the fixed resistance of the resistor, and the bias voltage across the resistor will increase accordingly, thereby reducing the strain on the tube. So yeah, even though 'heat is the enemy of electronics', running a cathode bias tube at Pmax is 'safe' and will produce a better sound all-in-all than running the tube at a colder bias point. YMMV. I say it's okay. The next guy will come up with some other hype. Why not try it and see?

[mresistor]

https://edcorusa.com/products/xpwr257-380v190-0-190-120ma-6-3v-4a?_pos=1&_fid=3600cf9a8&_ss=c    solid state rectification might be necessary


aha   here ya go  https://www.hammfg.com/files/parts/pdf/269EX.pdf

[tubenit]

One of the two 6BM8 amps I built used these Edcor trannies.


with respect, Tubenit

[HotBluePlates]

Now, if only someone could tell me how the Champ and the Princeton Reverb can both work using the same PT, I'd be a happy solder jerk.  One is a simple two-tube amp, the other is a six-tube complex amp with effects.  How can they both be happy running the same PT?

PRR told you the truth:

... If one part serves two models, you buy for both and are less likely to run out. ... So at some point Leo the bookkeeper realized that buying all the same part was the best business. ...

@PRR,
Thanks for your comments on the Champ/Princeton question.  I'd already answered my own question by comparing the amps' schematics, in reply #16.  They did it by using various combinations of rectifier tubes and OT winding resistance to get the various B+ levels out of one PT model. ...

Not quite, because there's a fatal flaw in the logic:  you can't trust the numbers on the schematic that much.  Bottom line, Fender sized the 125P1B transformer for the Princeton amps, then let the chips fall where they may in the Champ & Vibro Champ amps.

I used to own a 1965 Vibro Champ, and thoroughly measured voltages.

   - It took a "wall voltage" of 106.9vac to get 6.3vac on the heater wiring.
   - When reduced in this manner, I got 323-0-323 (not the 315-0-315v noted on the schematic).
   - 1st Filter Cap voltage was 367vdc (schematic says 355v)
   - 6V6 plate voltage was 366v (vs 342v on the schematic)
   - 6V6 cathode voltage was 17.95v (with a 465Ω cathode resistor)

Note that my figures don't match the schematic.  And the Vibro Champ schematic doesn't match the figures on the Champ schematic, either.

Despite how good Fender's documentation might be, it's only there as a rough guide.  There are even changes in parts (especially transformer characteristics) year-to-year in vintage Fender amps that aren't documented anywhere.  Changing PT voltages happened a lot.

[Lectroid]

@mresistor, @tubenit,
Thanks for these Edcor transformers picks. The Hammond has the voltage but does it supply enough current at 75mA?  Anyway either of the Edcors would work.

@HotBluePlates,
Thanks for the correction.  I just know all manufacturers cut costs wherever possible--my actual knowledge of Fender practices is minimal.  I'm a hobby builder, never been a repair tech.   
PS:  I never question PRR's knowledge.
 
But seriously, I am grateful to everyone who chimed in, and the many solutions and scenarios they presented.  This thread has given a huge boost to my understanding of power tube biasing and general power supply philosophies.  You are all the best.