Newbie Question: Load Resistance

[aumamps]

Greetings,
Please go easy on me. I'm here to learn/clarify a few things

When learning how to design amplifiers, I'm running into alot of a chicken and the egg/catch 22 scenario.
(In order for me to design/tweak a circuit, I need to have an idea of the power supply but in order to design a solid power supply I need to know the circuit)
I understand that the power supply is the heart of the circuit, so I am learning the math as I have a pile of transformers I'd like to use to build some amps...however when designing a power supply, there is the Load Resistance at the end of the equation. This is where I'm having rouble getting any real go about way of getting this figure in any of the design books that I have purchased.

I understand that this is the cumulative resistance of all of the circuits in the amp, but how do I really know the resistance if I don't know my power supply? If you understand what I'm asking.

How can you/do you "quickly" sum up an entire circuits load resistance? Am I just adding up the anode resistances at a certain voltage according to the data sheets of the tubes used?

Would someone be kind enough to do this for me with a simple circuit such as a Champ or a Deluxe as a reference?

Thank You

[tubeswell]

With a Power Transformer, the load is important to know from the perspective of the power rating of the transformer. A bigger load will demand more power. It is important have have a PT with a sufficient power rating for the job you intend it to do, otherwise you can fry the PT.

The load on a PT is formed by the sum of all the current pathways between each secondary winding and the ground return.

To work out the mains fuse requirement, you need to sum the volt.amps draw of all secondary windings, and then divide the total by the mains voltage (on the primary winding) to get the total current draw on the primary at the mains supply. This will guide you as to the mains fuse size you need (which typically is rated about twice the primary current draw). To work out the likely VA for each secondary, you need to consult the design of the amp you intend to use the PT in. For DC-resistance pathways (such as bleeder resistor networks, voltage dividers between source and return, and heater winding hum-balance resistors), use Ohms law to calculate current based on known resistances and expected power supply voltages: Voltage/Resistance = Current. For tubes, the tube current of each tube will tell you how much each tube draws from the HT winding and the heater winding. Sum the total of all the tubes for each respective winding. Same for any tube rectifier you might use)

Not all power supplies are the same in terms of load regulation. The load regulation will determine how much the power supply sags under a given amount of load. More sag per given load = poorer regulation. Certain types of rectification methods can also detract from load regulation. If you want a sturdy power supply (important for powerful clean sound), you don't want much sag.

So if you have a bunch of PTs that you don't know the load regulation of, you can try making your own assessment of each one's load regulation by comparing load/no-load voltages with a 'dummy load'. For the most telling results, use lowish resistance (but sturdy) wire-wound resistors across each secondary winding to simulate higher load. Do the voltage measurements (carefully) with the primary side plugged into a mains voltage supply (with an appropriate mains plug, mains fuse and switch, and current limiting/measuring device). Measuring how much the voltage on each loaded secondary winding sags, compared to a no-load situation, will give an indication of how each PT is going to perform under load.

The topic of load resistance for output transformers (OTs) is a different matter.

[Tony Bones]

aumamps, PS design can be an exact science, but when using salvaged parts, then rough estimates and cut-and-try get results quickly that are good enough.

It's usually easier to think in terms of total current draw than load resistance. Just estimate the current you expect each stage draw given the B+ voltage you expect from that rusty transformer. To guess at final voltage, hook up the primary to wall voltage then measure the open terminal secondary voltages. Maybe multiply by 0.9 to estimate the VAC when loaded. Rectifier datasheets often have graphs that show what the final VDC will be with different loads.

If you really want to calculate load resistance, then use ohms law: R = V/I

Some people enjoy playing around with Duncan's PSUD2.

It's often an iterative process and getting to within 10% is good enough!

[martin manning]

I understand that this is the cumulative resistance of all of the circuits in the amp, but how do I really know the resistance if I don't know my power supply? If you understand what I'm asking.

Make some assumptions and run the numbers. If your assumptions need revising, adjust them and repeat. If you are reproducing a known circuit, all this has been done before, and you have some useful information to get started.

For reusing salvaged transformers, I agree a first step would be to measure the unloaded secondary voltages and see if you have something that suits your intended purpose. Determining the current capability of each secondary requires loaded measurements. You could take whatever current is required to reduce the voltage by 6-8% as the limit. Note you will need some large resistors for this. A typical 330-0-330, 200mA HV secondary would require 330/0.2=1650Ω and dissipate 0.2x330=66W.

How can you/do you "quickly" sum up an entire circuits load resistance? Am I just adding up the anode resistances at a certain voltage according to the data sheets of the tubes used?

No, that would not be correct, but you are on the right track in adding up the contribution of each stage. It's easier to work with current, which you can determine from a load line analysis. You could make some rough estimates, though. For example 1 mA for each 12AX7, and for each power tube the (idle) current could be estimated as the assumed plate dissipation divided by assumed plate voltage.

[aumamps]

Thanks guys! Your time is much appreciated.

I have some transformers from an old Baldwin 54A and I know what their ratings are. The PT is a 300RMS - CT and was thinking of a Rocket clone but 1) I've done enough clones and 2 that voltage is a little too high for the circuit. So I've been debating about the best way to drop the B+ (mosfet versus Choke input) but also just debating how to tweak the circuit in several ways besides just changing out components like a monkey. Love the tone of the Rocket though. That's the tone I'm looking for.
Also have a few other smaller transformers I could use to get my feet more than wet and really play around with all of the different tubes I have to design something unique and maybe just stick with a rocket clone for gigging purposes since I need an amp for live work soon

[Phil_S]

You should be able to search for a thread that I started about choke input. I tested a few chokes and there is some information about the specs and results. It might be helpful.

As far as using that Baldwin PT, if you are lucky enough to know the tube compliment of the donor amp, you will be far ahead of the curve by knowing that you can use it for an amp with the same basic requirements. How far beyond you can go in terms of mA capacity can be tested for the way Martin says. Unfortunately, the only sure way to know -- unless you get lucky and find the manufacturer's design spec -- is to let the smoke out and you don't want to do that. So, it is best to test for the circuit you intend to build and quit there.

[aumamps]

I don't know the mA rating of the PT but it powered 6 El84s and about 4 12AX7s so I know its definitely capable of a standard VoxAC30 or Marshall 2*EL34 power amp

[Phil_S]

That is good information. You know for sure it can support that tube compliment. You can look up the information on the tube data sheets and then make an educated guess. I suggest you've got ~350mA on tap. The data sheet for the EL84 says in class AB1 push pull fixed bias, Ia max= 92mA, which I assume is for a pair. Ig2 max (screen) is 22mA. Allow 1 or 2mA x4 12AX7's. Almost more important is knowing how much heater current you have. That tube compliment draws 5.76m @ 6.3V. I'd be careful not to push heaters. One thing you can easily do is use 6V6's, which draw much less heater current and are generally voltage compatible with EL84 circuits. I'd guess a pair of 6L6 or EL34 will be fine with this PT.

If I were load testing it, I'd start with that information and see where it goes. You didn't say what the unloaded HT secondary voltage is. I'll guess at 250-0-250. We know Ohm's law says V = I * R. We know V and I (approx). Solve for R = V/I. 250V/350mA = 714Ω. Start testing with a load of 1KΩ that is rated for the proper number of watts. In this case, P = I * V; 350mA * 250V = 75W. I'd choose a higher rating, probably more than100W rated. Decrease the dummy load until you get a ~5%-6% drop from 250V. I don't think I'd push my luck too much. I'd quit when the load produces 235V. At that point, you know the R value of the load and the voltage. Solve for I. That's likely a safe number for maximum current.

You will know the actual unloaded voltage, so remember recalculate everything.

When you do this, keep one hand on the off switch. If you are going to smoke the transformer, this happens very, very rapidly. (Don't ask how I know this.) At the first hint of any burning odor, be very quick to turn it off. Err on the side of caution.

When building the load, multiple resistors can be bundled to achieve the wattage rating. 4x 25W 4K resistors in parallel provide a 1K 100W load.

If you are not imagining this, it is deceptively simple. Screw everything down so nothing is loose. I use a piece of scrap wood and make sure anything carrying electricity dosen't touch the wood. (Heat = fire.) Get some terminal strips. I like the screw down kind so I can swap resistors. Run one B+ supply into the one end of the load and the other B+ supply into the other end. Measure the voltage across the load and monitor it.

[aumamps]

Thank You Phil S, that's basically how I've been doing, or assuming, the math at this point. I do a calculation based off of expected voltages on the data sheets and it's corresponding max current (as well as an average between max current and idle current).
The confusion for me is that in the dozen tube amplifier design and educational.books I have, this is never made in direct lamens terms, especially when in some of these books the power supply stage is given little attention and is usually a later chapter in the book, which seems contrary to me as there is a lot of shaping of the sound (not necessarily the tone) that can be done with proper filtering, specifically the bottom/low end.

I know people will shake their heads at that, but I am a self taught learner in most things in life and considering tube amplification is a specialty of the past, I guess the authors sort of take it for granted this basic math, yet this is not a procedure you'd generally find in a modern basic electronics textbook

[Tony Bones]

I'm not shaking my head; I think you're right that power supplies are taken for granted. There's a lot more to them than a single chapter can cover, no matter if it's as the beginning or end of a book.

On the other hand, if you're building from salvaged parts like you are, there's really not much you can do but build it and see what you get. That's OK and can even be a fun adventure. What's the worst thing that can happen? You learn something....

[pompeiisneaks]

I'm not shaking my head; I think you're right that power supplies are taken for granted. There's a lot more to them than a single chapter can cover, no matter if it's as the beginning or end of a book.

On the other hand, if you're building from salvaged parts like you are, there's really not much you can do but build it and see what you get. That's OK and can even be a fun adventure. What's the worst thing that can happen? You learn something....

Yeah Merlin used to have a power supply book but he stopped selling it because he felt it had too many errors. He's hinted that hey 'may' do a 2E but he has been silent since. He did put a decent amount of the power supply stuff in the two other books as well, but still, like you said, I think a single chapter (or two in two books) is still light reading on how much complexity needs to go into power supply design.

~Phil