Regulated Voltage Supplies

[bob p]

There is also a trick I saw in an old Ham radio manual that put the choke in the negative leg. Works just fine for a choke input filter.
Tom

Got a question about that Tom -- if you put the chokes in the grounded side of the filter and the transformer case is also grounded, won't capacitance between the core and the secondary winding shunt the chokes and bypass some of the ripple around them, thereby reducing the filtering? Since the negative terminals of the caps are typically the ones that are grounded, the chokes are normally placed in the positive side of the filter to prevent this phenomenon.

Thanks again to everyone for their help.

[kg]

I have a question about running the chokes in parallel: Do the potential problems of running PTs in parallel also apply to running chokes in parallel?

not to nearly the same extent.

what you want to eliminate obviously is current hogging, where one parallel device is carrying a majority of the load compared to another.

in the case of the power tranny it's a bit more delicate because the ideal PT is a voltage source. the closer one comes to that ideal the more critical exactly matching the secondary windings becomes.

in the case of a choke, though, you're not going to have the issue of circulating currents, because the choke isn't SUPPLYING anything per se.

of course, the thing that helps you out with current balancing in any situation is DCR... the more there is, the more likely the current will be shared equally.

Hammond specs their chokes as having 15% tolerance for L and Z. To be on the safe side, would I have to assure that I get chokes that are matched for Z?

well it would be ideal to get them as closely matched as possible. when i picked up my first pair (i've got 2 pairs! ) of 1650w OPTs they were consecutively serial numbered, meaning they rolled off the assembly line one after another and would be as closely matched as physically possible without manually measuring and matching one by one. so see if you can get serially numbered chokes from whomever you source them from.

if it makes you feel better you could match perfectly for DCR at least by adding a low R ballast resistor once you get the chokes in circuit. since you're talking about pretty low inductances to begin with there are probably not a lot of turns on the core and i'd be surprised if there was a big variation in DCR or in L.

[bob p]

well it would be ideal to get them as closely matched as possible. when i picked up my first pair (i've got 2 pairs! ) of 1650w OPTs they were consecutively serial numbered, meaning they rolled off the assembly line one after another and would be as closely matched as physically possible without manually measuring and matching one by one. so see if you can get serially numbered chokes from whomever you source them from.

Interesting that you've got 2 pair of 1650w. Is it safe for me to assume that the second pair indicates that you like them? Or did you buy them all at once? What kind of amp did you build for them???

FWIW, I've been thinking about the 1650T for this project. Same Z as the 1650W, at 120W its less than 1/2 the power handling of the 1650W, and at 14 lb., its also half the weight!

[kg]

Interesting that you've got 2 pair of 1650w. Is it safe for me to assume that the second pair indicates that you like them? Or did you buy them all at once? What kind of amp did you build for them???

well they're basically one of the biggest off the shelf trannies you're going to find for "reasonable" money. the first pair went into the BAGA of course nearly 10 years ago now:




then, a few years later, while still bitten by the big ass tube amp bug, i started collecting a second set of parts: for plate supplies, the large 1k5va industrial EI tranny, the DIY 1kva toroidal tranny. i've got a couple of 6v3ac filament trannies, some interstage trannies, and sockets and tubes (el509). first it was going to be a backup big ass guitar amplifier, then it was going to be a big ass bass amplifier, then a big ass hifi amplifier. the parts still mock me in my basement.

all i need now is time and motivation and speakers!

[bob p]

ah, the BAGA. i should have connected the dots.

[R.G.]

So Bob - did the objectives ever settle down to to the point where we can design?

[Steve Conner]

Design? Why spoil the fun? ;-)

[R.G.]

The only joke I can remember that was of Russian origin was this:

A woman and a man heretofore unknown to her are in the same compartment in a train on a trip across Siberia.

On the first day about noon the man smiles and says "Hello." The woman says nothing.

On the second day at about noon, the woman looks him in the eyes and says "Hello." The man says nothing.

On the third day, about noon, the man says to her "Enough of this foreplay. Take off your clothes."

[bob p]

Hi R.G. -- since you haven't posted for a page or two, I thought we might have lost you.

Did you ever get a chance to double-check my numbers? I'd really prefer to know that they're reasonable rough estimates of the two possible designs, as knowing that the data is good would really help to guide my decision.

FWIW, I've got a 2KVA 240/480-120/240 single phase transformer coming my way, so I guess our design choices will be limited by what we can build with that!

[bob p]

i thought i should mention an error in one of my previous calculations. when I was spec'ing chokes for an LC filter, I read the value for DC output voltage from the wrong column of the nomogram in Crowhurst & Cooper. What I had thought was the DC voltage was actually Imax. Oops.

here's what I originally posted in Post # 50:

Let's look at the choke idea and crunch some numbers. Because a Class A1 amp doesn't have very significant current swing, the choke calculations are pretty easy. Since Imin and Imax are almost exactly the same for a Class A1 amp, Iopt = Imin = Imax = 1.4A. Now all we need is the correct value for L at a 400VDC operating voltage. Looking at the nomograms in Crowhurst & Cooper, here's how the numbers work out:

Imin  = 1.4 A
Imax  = 1.4 A
Iopt  = 1.4 A
MaxDC = 400 V
Lopt  =  30 mH

The actual Lopt value from the nomogram is about 450-500 mH. What a coincidence. The tradtional formula for critical inductance gives Ken's result of 300mH. I think that the nomogram provides numbers that are about 25% - 30% over the critical inductance values to be conservative. So it looks like something in the range of 500 mH should do it. As Ken previously mentioned, to satisfy both the mH rating and the current rating, it looks like multiple chokes would be needed at each filter stage.

Another option would be to just create parallel supply rails (one for each channel) coming off of the PT secondary. I'm thinking that the best way to do this would be to have parallel supplies coming off of the transformer, with each channel's rectifier circuit including its own bridge, choke, cap, etc. Doing this effectively cuts the Imax requirement in half, doubles Lopt, and makes the right chokes easier to find.

For one channel the Crowhurst/Coooper nomogram tells us:

Imin  = 0.7 A
Imax  = 0.7 A
Iopt  = 0.7 A
MaxDC = 400 V
Lopt  =  1 H

That makes choke selection a bit easier: Hammond 193S, 1.0 H, 1000 mA, 3.50 R, 1000 VDC, 4.5 lb.

So I'm thinking that each channel would attach to the BFTransformer, and the supply would look like this:

       --- bridge - L - C - L - C
 PT---|
       --- bridge - L - C - L - C

and the chokes would be Hammond 193S and the caps would be some big value like 470uF.

With those numbers, and Steve's approximation for the internal Z of the transformer, the models suggest that this setup might idle in class A near the target of 400V. As long as we're close, load lines should tell us the rest. Which reminds me:

Has anyone got a diagram of composite loadlines (normal+inverted) for designing a PP 6550 output stage? I'm just wondering if this kind of chart is already available, so I don't have to re-invent the wheel.

thanks!

[kg]

Has anyone got a diagram of composite loadlines (normal+inverted) for designing a PP 6550 output stage? I'm just wondering if this kind of chart is already available, so I don't have to re-invent the wheel.

it would be nice if this type of thing were available, but the problem is that the composite loadlines are dependent on what you set your idle current to be.

it's a bummer! "properly" doing the PP load line is a PITA process indeed.

[bob p]

yeah, i was afraid of that, but i thought it wouldn't hurt to ask.

obviously, the easiest way to build your amp is to find iron that will allow you to whip together something like the default operating conditions that are listed in the tube data sheets. that's one reason that i've leaned toward the default setups that are listed in the data sheets -- if you can find the iron to reproduce those conditions, you don't have to manually create the composite loadlines. as long as you can get pretty close, you're OK. its when you're pretty far off that you have to gothrough the trouble of creating your own composite loadline charts.

there's got to be an easier way to do this. i'm thinking that by now somebody has to have done this sort of thing with Spice or some other tube modeling software.

is anyone familiar with tubecad? i took a brief look at their web site, and i got the impression that it was geared toward triodes. i honestly don't know if that's true or not. i'm hoping that the software would cover pentodes as well.

is doing this sort of thing even possible with software? if so, the convenience factor alone would make the software money well spent.

[bob p]

bump.

there's got to be somebody who's familiar with tubecad. ray?

i'm thinking that somebody has to have drawn composite loadlines for 6550s running in Class A at the spec values on the GE data sheets. nobody? i don't mind doing it with The Gimp, but i'd rather not re-invent the wheel if i don't have to.

[bob p]

Well, the Big-Ass 2 KVA transformer arrived today. At 38 lb., its a heavy lift!

R.G., you were right about the need for a modular design for this amp project!!!



Side Note: I would NOT recommend purchasing anything from 1MachinerySales.com, either on their web site or on eBay. See Below.

The PT that I selected for the job is a used Acme T-2-52012-S. Its a 280x480 primary, 120/240V secondary single-phase transformer that's rated at 2 KVA. I'm running it backwards, using a 120 VAC input to produce a 480 VAC output.

The first thing that I did was to rewire the transformer for a 120 VAC "secondary". It had been previously wired for 480 to 240 VAC. Then I took it down to the bench to check it out.

I had expected that with 38 pounds of iron, the inrush current might trip a 20A circuit breaker when I tried to energize the coils. I wasn't disappointed! Even though I tapped into a 120 VAC outlet on its own circuit, located right next to the service panel, every time that I tried to power up the transformer, it tripped the circuit breaker.

After a few failed attempts, I fell back to Plan B -- plugging the transformer into a variac, setting the variac to 0, turning it on, and dialing it up from Zero to 120 VAC on the primary side. That worked like a charm, and the PT is putting out 480 VAC with no load. Best of all, its also quiet with no load!

So this has got me wondering about inrush limiting in the amp -- should I rewire my "primary" to operate at 240 VAC instead of 120 VAC, or am I better off using 120 VAC and some sort of dedicated inrush limiting device in the amp? For practical purposes, I don't want to power up the amp on a variac.


Side Note: I would NOT recommend purchasing anything from 1MachinerySales.com.

I purchased my PT using PayPal and a credit card over two weeks ago, and never received any confirmation from the company. I contacted them by email every day for two weeks asking for the tracking information, and all of my e-mails went unanswered. I called their customer support staff on the phone, and nobody knew how to give any sort of answer once they had taken my money.

Finally, after 2 weeks of no reply and no merchandise, I asked VISA to issue a chargeback for non-delivery of goods. Then of course, the transformer arrived by FedEx the following day, and the merchant started responding to my emails by sending me the tracking number AFTER the merchandise had already been delivered!

Worst of all, the transformer was poorly packaged, and the mounting tabs were all bent in transit to the point that they were not useable. I had to spend 15 minutes with a hammer and a vise bending them back into shape.

All things considered, this merchant SUCKS. They take your money, they pack inappropriately, they refuse to offer the tracking information, and they refuse to communicate with you until you get their attention with a Visa chargeback.

[R.G.]

That's some monster chunk of iron and copper!

I think you might be able to power it up with an inrush thermistor or a time delay relay. The thermistor is the best option if you can finesse the design to limit the current but still go hot and conductive at 'way below full load current. Otherwise, a power on relay made from a solid state relay or two should help. I can help with that as I've faced this issue before too.

I like the "quiet" too. I was afraid they might work the iron so hard that you got a lot of no-load 180Hz core flex.

So - we ready to design this up?