PT question

[bigdaddy]

Since I have been researching very low wattage amps I find some issues trying to acquire the right power transformers. Not to mention OTs.

My question is rather simple. For discussion sake let's say we have a typical overbuilt Champ amp PT. This is too much voltage and amperage for these small amps. They are usually around 320-0-320 and 120+ma or more.

Why can't I just use the center tap to ground and only one leg of the secondary? That would give me 320 volts. Use a diode grounded and another one for the B+ rail. I have schematics with something like this on the PT.

Or can I use one side of the secondary and the center tap together with a bridge rectifier with no center tap? if I did these things what voltages would I see?

[HotBluePlates]

Why can't I just use the center tap to ground and only one leg of the secondary? That would give me 320 volts.

But it gives 320vac.

The normal PT and rectifier circuit is a center-tapped secondary and a full-wave rectifier. When the PT is energized, one rectifier plate, then the other, conducts alternately. Each plate sees the voltage from its side of the secondary to the center-tap, and because the to halves of the rectifier conduct alternately, you get 120Hz ripple to the filter. In this case, the ac input to the rectifier is still 320vac.

So what you propose will give the same voltage to the rectifier, but (harder to filter) 60Hz ripple due to being a half-wave rectifier. Or you could use a full-wave bridge on half the secondary, but you still get 320vac input, now with 120Hz ripple output. Due to drawing from half the secondary, you get the same voltage output at half the current rating of the full winding.

You really want to look for low voltage power transformers. You'll probably have to accept having a seperate filament transformer. 120vac or 140vac seem like they should be available, even if only as isolation transformers.

[kagliostro]

This isn't a solution

only a proposal

what about to put a (little enough small) transformer connected between the CT and ground as to decrease the voltage

a 20v in opposite phase will decrease the B+ rail by the same voltage

that is my idea about to don't change the PT adding only a small enough one to reduce voltage

(if the PT has 120mA but the circuit necessity is only for 70mA, you can use a 70mA 20V transformer)

I would like to here about this idea from the more expert friends of the forum

Kagliostro

[mresistor]

Not and answer to your ? but does this place have anything you could use?   http://www.musicalpowersupplies.com/4.html

also wouldn't Hammond Mfg. have a suitable HV PT?

[jjasilli]

Antique carries Hammond PT's: 190-0-190 65mA (no 5V); and some low mA units @ 250-0-250; and 275-0-275;  after that, mA ratings skyrocket.  Maybe Edcor will build you something custom.  Or keep an eye on eBay for some vintage unit that might work.

Seems you only need 30 - 50mA if you want to under-rate it.  The stock Champ PT is 70mA which does seem excessive.  But maybe Leo knew what he was doing with that "over-rated" PT; or found out the hard way. 

[andrew_k]

Cheap plastic portable reel-to-reels are a great source of 250-0-250 or 250-0 PTs with 6.3V winding(s).
Not that I'd cannibalize other gear to feed my hobby... never..   

[bigdaddy]

I just got a 250-0-250 Hammond to mess with. I am going to make a BF champ using both triodes from 12A_7 type tube in parallel. I'll see what the voltage is with a 5Y3 and take it from there.

The new Hammond transformers have a new thing, at least on this one. There is a 115v and 120v Primary, Gray/white and Black/white.

I'll be using a fender reverb tranny for output. Some say it lacks bass, I say good, I don't want much of anything just to make a good tone at an extremely low volume..

[HotBluePlates]

You could "tune" the voltage by using lower-value filter caps, regardless of the current rating of the PT.

We always seem to go "bigger is better" on filter caps to nuke hum. I don't recall the exact number, but if you had no filter cap and replaced the amp with a resistor, the measured d.c. output from the rectifier would be quite low. That's because the output from the rectifier is a pulsating d.c., so the average output is low (IIRC it's somewhere around 0.38 * peak a.c. rectifier output).

The filter cap stores energy to hold the voltage up between charging impulses. If the load drew no current and you used a filter cap, the output voltage of the rectifier would equal the peak a.c. rectifier output. So we'll land somewhere between the extremes.

So if we use small caps typical of vintage amps (or even smaller), we could reduce the resulting d.c. voltage at the expense of more ripple. This would be easier to workaround in a push-pull amp, but will work either way. Use a small cap (8-16uF) as the first filter after the rectifier. Follow that up with a choke or resistor; remember that the choke must be rated for full amp current and the resistor for amp current and resulting dissipation. You can use bigger filter caps after the choke/resistor to knock down ripple as the voltage conversion has already taken place.

Ripple is not a big deal at the output tube plate, but is more critical at the screen and preamp supply points. Under this scheme, the screen is the 3rd (or more) filter cap in the power supply. This method is often required to get low hum in SE amps using full-range (guitar range, anyway) speakers with response near 60Hz.

Anyway, it requires more room for the power supply, and the willingness to experiment with mutliple values to see what really works for you. There are charts out there for rectifier output and current draw, but that's a whole new thread with a lot to post and consider. This is a technique you might be able to apply to an existing amp to try out.

[John]

Not to hijack, but HPB... if I understand you correctly, you can lower B+ by using a lower value cap coming off the rectifier tube?

[jojokeo]

B-daddy,
Here's another link for small powered Hammond trannies w/ the best prices I've found if you don't want to use a torroid as I suggested on the other thread - there's only so many options available (which is plenty really unless you custom order something but that's more dinero and not worth the time).
http://angela.com/hammondlowpowerbiaspowertransformer.aspx

If you want to dump B+ w/out hassle or craziness just install an appropriately rated power resistor after the rectifier and filter as normal - this is the easiest and cheapest way IMHO. The next thing I'd consider is the zener diodes to ground off the CT ground. (I'm kind of waiting for PRR to say something such as "as long as you're w/in reason or 20% of a tube's ratings you'll be fine and don't worry about it".) 

[John]

If you want to dump B+ w/out hassle or craziness just install an appropriately rated power resistor after the rectifier and filter as normal - this is the easiest and cheapest way IMHO.

I was thinking that, but I read so much that sometimes I get confused. ('magine that)

[darryl]

If you want to dump B+ w/out hassle or craziness just install an appropriately rated power resistor after the rectifier and filter as normal...

This works in a single-ended amp, which is always Class A, but a push-pull amp is often Class AB1. A dropping resistor in the power supply of a Class AB1 amp will produce sag under load ( which may or may not be a bad thing  ).

The next thing I'd consider is the zener diodes to ground off the CT ground.

A zener diode ( or several in series to share the power dissipation ) can also be used in the B+ supply, in place of the resistor described above. This will produce a fixed voltage drop, regardless of fluctuations in the current drawn under load.

[John]

Got it. As always, depends on what you're looking for. So changing cap value will lower voltage without the "sag". Or, I suppose you could do the resistor parallel with another cap?

Anyway, I am hijacking the thread, and didn't mean to. Carry on! (I'll keep reading)

[bigdaddy]

I have been doing that lately, using a 10-20MFD cap for the first cap and then a 47MFD and 22 or even 10s after that. I actually built two similar amps to test various things side by side to see the results of different things like that. Now I am on another road......

My first attempt at a low watt amp will be a BF champ with a 6AK6 since it's closer to what I am used to working with. I am not used to dual triodes as power tubes. I read up and they can take 300v plates and 250v screen, so I think it's best to keep it around 250v and a little more for the plate. This will make things easier. I am going to divide the B+, one side for the preamp and one for the power and see how that does, kind of like decoupling them I guess. Maybe I use a pair of diodes too. Go from the B+, switch, 10mfd cap with 220K bleeder then split with diodes and use the proper dropping resistors and filter caps.

I also head the Gilmore 1/2 watt amp with full PI, it screams but the head is like $1000....YIKES.

[HotBluePlates]

Not to hijack, but HPB... if I understand you correctly, you can lower B+ by using a lower value cap coming off the rectifier tube?

The voltage at the PT secondary is a.c. If you have a grounded CT, then then voltage is referenced to ground and both rises about 0v to a positive maximum, and goes below 0v to a negative maximum. The rectifier lops off the negative portion, and only conducts during the positive portion.

So the natural output of the rectifier is constantly moving from 0v to some maximum. If you know the RMS voltage, you can calculate the peak voltage of a sine wave (or half-sine in this case). The peak is 1.414 times the RMS voltage; if the PT voltage is 320vac (or 320v RMS), and the rectifier has zero loss, the output is moving from 0v to 320 * 1.414 = 452v

If you connect a cap to the output of the rectifier, and you draw zero current, the cap will charge to 452vdc and stay there. As soon as you draw current, the voltage might drop, and usually does. If the current drawn is due to a short-circuit, obviously the voltage must drop to 0v.

The ability of a capacitor to maintain the voltage level following the rectifier is a function of its capacitance. The bigger the cap, the more energy it can store and release over time. If the cap is small, it will result in a lower output voltage than a bigger cap. How much lower is question answered with a series of steps that few of us bother to take, and that would have to be shown at a later time. Or read the rectifier section of the Radiotron Designer's Handbook, Vol 4, or a couple other old engineering books.

I don't suggest it as a catch-all way to drop supply voltage. But a number of suggestions to obtain a low supply voltage in a new build were given and set aside in a search by Bigdaddy to make a low-volume amp. We had already suggested choosing a smallish PT, using the VVR circuit (or powerscaling or similar) to drop voltage, applying a bucking voltage to a PT primary or secondary, etc. This was a last-ditch suggestion to fine-tune a supply that is otherwise close to the desired voltage but too high.

[John]

Thanks for the explanation!

[jjasilli]

Talk about hijacking:  Thanks to Hotblue for triggering in my mind a solution to the high voltage in my Bogen CHB-20.  The Bogen CHB-20 uses a voltage doubler to produce 400 plate volts on PP power tubes -- el-84 like substitutes rated for 300V.  One alternative to drop voltage is a string of zeners.  But this seems in-elegant.  And zeners cause voltage regulation which is not my cup of tea in a guitar amp. 

The amp's voltage doubler uses 2X SS diodes with 2X 40mF filter caps.  Solution using the Duncan PS simulator:  drop the doubler's cap values to 4uF or 8uF > 330Ω Resistor > 20uF or 40uF filter cap > plates.  Result: ~350 -360V = Trainwreck territory.  Seems good to me.

[HotBluePlates]

Give it a shot, test and adjust. Minor changes in current will probably make a fairly big difference when you're talking about a voltage-doubler supply and an output section that actually delivers some power.

Be prepared to be wrong, but give it a shot and tell us what you come up with.

[RicharD]

I'm a little late to this thread, and honestly I didn't read it all whereas I am confused at to roughly how much voltage is needed to be sluffed off.  A trick I've used to drop 10 or 20 volts is to slap a resistor and cap between the CT and ground.  This puts the CT at a negative potential.  Depending on the total current, you could easily drop more volts.  My example of 20V was based on an experiment @167mA using a 10W 120R resistor.  Example attached.

[John]

Yet another file in my Amp Stuff folder. :) Thanks!

[eleventeen]

"One alternative to drop voltage is a string of zeners.  But this seems in-elegant.  And zeners cause voltage regulation which is not my cup of tea in a guitar amp. "

I believe HBT was talking about inserting a string of series-connected zeners in the B+ leg to dump 100 volts from your too-hot power supply. NOT connected from the B+ to ground (using a current limiting resistor, thank you very much, in the conventional way we think of using zener to regulate power supplies)

http://cgi.ebay.com/1N4733-5-1v-1-watt-Zener-100pcs-/400059706974?cmd=ViewItem&pt=LH_DefaultDomain_0&hash=item5d256aae5e

Here you can buy 20 qty 5.1 volt 1 watt zeners and have 'em shipped to you all for $6.50! Just remember to get the $17 worth of terminal strips to properly mount 'em, LOL!  

But, their effect would not be to "regulate" the supply per se if used in this fashion. Twenty qty of them would indeed create a regulated 102 volt voltage drop between the highest B+ point they linked to and the lowest anode (the base of the triangle in a zener diode symbol) in the series string. But a meter placed at that final diode (or anywhere else) would not show any better supply regulation than what's present at the normal FW or half wave supply itself. Current capacity = ~~200 ma. Be aware that the zener only "zeners" if installed "backwards" to how we normally think of diode conduction. Otherwise, it will just act as a regular diode and drop .7 * 20 = 14 volts. In other words, the zeners would be hooked up in that "backwards" way we see diodes hooked up in a typical bias supply.

Hey, what's inelegant about 20 series connected diodes?? Just kidding. You could use a smaller number of higher voltage-drop zeners, but zeners aren't known for high current capability. You'd more than likely immediately smoke a 100 volt zener if you tried to use it in this fashion.