Making a Dummy Load

Morden2004

Relaxin'
One of the items on my DIY list was to make a dummy load for testing and troubleshooting amplifiers and receivers.

I was expecting to have to cobble together a bunch of 50 watt resistors =- having done this several years ago.

The reason for expecting to use 50 watt resistors was the cost - high wattage resistors are expensive and I'm on a budget.

But! eBay to the rescue. I love that marketplace.

I located a guy selling a batch (15) power resistors - mostly 50 ohm 100 watt which I was quite prepared to lash together to give me 12 ohms for testing with.

Well, this guy also had what is in the picture below - a pair of 8 ohm 300 watt Milwaukee ballast resistors. 300 watts!!! Exactly 8 ohms!!!

My Dummy Load will be ready shortly. The big (12 inches long!) double resistor at the top of the picture is the 8 ohm 300 watt unit. One of the 50 ohm 100 watt resistors is showing in the box.

Paul
 
Not a good idea. Wire wound resistors will present different loads at different frequencies.

X
 
Punker X said: Not a good idea. Wire wound resistors will present different loads at different frequencies.
Not being smart mouthed here - but so do speakers & x-overs. An 8 ohm rating is, for example, 'nominal' and usually measured at 1 Khz. The actual 'impediance' could run anywhere from 5 or 6 ohms to 20 or 30 ohms depending upon the frequency.

I know that wire-wound resistors can present a somewhat different looking load to an amp. In fact, I believe they are less stressful - little or no reverse current. I'll have to ask EchoWars for clarification on that.

Most wire-wound resistors are not truly inductive (no ferrous core) so they presenty mostly a resistive load. The only risk is if they don't hold their nominal value when hot.

Paul
 
Punker is just as knowledgable about this stuff (or more so) than I.

Still, these are probably OK at frequencies of about 5K and under. At 1KHz, where you would probably want to check for full output, I'd bet they are peachy. At 10KHz and up, I'd be skeptical.

Parts Express sells 4 and 8 ohm 100W audio dummy loads, and 4 and 8 ohm 200W dummy loads. Handy, and not too expensive.

http://www.partsexpress.com/pe/pshowdetl.cfm?&DID=7&Partnumber=019-030
 
No, I was sure he was correct in his statement that they can be problematic - but I was sure that it was usually at (quite ?) high frequencies and high output.

Thanks to both of you for your feedback. I will tread very carefully when I start using these - and I'll stay at 1 to 3 Khz just to be safe.

Do either of you know how you can tell if you are getting standing waves or reverse current (etc.)? If these resistors decide to induce chaos in my tests I would like a measurement method to monitor them. I don't recall ever seeing a method for measuring reverse induced current. I would assume it would be out of phase and at some fraction of the power input to the load. I have a scope but I'm not sure what in fact you would see.

I don't want to fry any vintage power transistors. I have been following Bolly's attempts to get some parts for his AU-D11.

Paul
 
The way to see what is going on is to look at the response with a square wave generator and a scope. Nasty overshoot and ringing would be a clue.

Still, for power testing up to about 5Khz, I'd think you are probably fine. Hell, it's probably fine all the way to 20Khz as long as you don't get crazy with the amplitude. In reality most testing is done at 1KHz anyway.
 
Thanks for the assurance. I'll be real careful just in case.

I want to get a chassis and mount that double-resistor securely and wire it up properly. I'm always reluctant to place speakers on an unknown amp.

I do have a good audio generator and I will experiment using square waves into my AU-999 as soon as this dummy load is wired. I'll post pictures.

Paul
 
I also think those resistors will work fine for your application. The best way to resolve if the inductive reactance will be a problem is to measure the resistors with a inductive meter. With that measurement, then you could calculate the actual impedence of these resistors at any audio frequency you wish.

As you said, most speakers are an inductive load and in fact I once saw an 'offical' RIAA speaker load (dummy load) model somewhere but of course I didn't save it and cannot recall the values.

Lefty
 
Hey, Lefty, I might do a Web search to see if I can dig up that page. That would be an good article to have for my project.

Thanks,

Paul
 
Originally posted by Morden2004
Hey, Lefty, I might do a Web search to see if I can dig up that page. That would be an good article to have for my project.

Thanks,

Paul

A quick goole search found these:

http://www.duncanamps.com/technical/dummyload.html

http://sound.westhost.com/vi.htm

http://www.amptone.com/g243.htm

I guess one should first define what they want out of their dummy load. If it is just to measure an amplifiers performance (power and FR) under just a pure resistive load or if it is to check out performance under 'real world' type speaker loads (check for stablity) without risking damage to a real speaker.

Good luck in your research, knowlege gained is always worth the effort.


Lefty
 
Last edited:
I've been hitting the books...

and after a 30+ year hiatus, I can't believe how much I have forgotten.

So, the topic is a dummy load for amplifier testing.

Here is a jumble of thoughts in no particular order;

1. Tube amps are typically tested using the 16 ohm impediance tap
2. SS amps are typically tested using the 8 or 4 ohm tap
3. Power output testing is usually done using 400Hz (manufacturers choice)
4. Power output testing is always done into a resistive load (reasons later)
5. Band pass and distortion measurements are typically done using an reactive load
6. Long speaker cable runs may eat up as much a 25% of your band pass power!

Well, back to the reason for using a resistive load for power testing. It's basic electronics but I must admit I have completely forgotten the voltage/current/phase relationship.

In a reactive load that includes inductance (like a speaker), the voltage peak always leads the current peak. What this means is that if you measure the power input to a reactive load you will need to compensate for the phase angle delay. Also, consider this; a complex resistive/inductive/capacitive load which includes the speakers, x-overs and the cable is basically a tuned circuit and therefore has current peaks based upon the tank circuit's resonance.

Hey, this is fun.

However, if no one is interested :D I'll keep this to myself (in other words - shut up, Paul)
 
I think I like Test 1, that sinewave generator/scope idea. I don't have a clue why you switch to squarewave in Test 2. BTW Lefty great link!

Greg
 
Bolly:

That square wave test is 'the' test used to measure bandpass (frequency response = low -> high @ a specific -dB point).

Because square waves are made up of (theoretically) an infinite number of harmonics above the root it presents a convenient frequency 'sweep' with which to test the subject amp's bandpass.

A 'perfect' amplifier (like, a Sansui BA-5000 :D - just threw that in in case B/F is watching) would reproduce the squarewave with a prefectly straight leading edge. However, in the real world that leading edge will be rounded off due to high frequency limitations in the amp - the more rounded, the less accurate it is at high frequencies.

:dunno: I think that's an accurate synopsis?

Paul :p:
 
Originally posted by Morden2004
Bolly:

That square wave test is 'the' test used to measure bandpass (frequency response = low -> high @ a specific -dB point).

Because square waves are made up of (theoretically) an infinite number of harmonics above the root it presents a convenient frequency 'sweep' with which to test the subject amp's bandpass.

A 'perfect' amplifier (like, a Sansui BA-5000 :D - just threw that in in case B/F is watching) would reproduce the squarewave with a prefectly straight leading edge. However, in the real world that leading edge will be rounded off due to high frequency limitations in the amp - the more rounded, the less accurate it is at high frequencies.

:dunno: I think that's an accurate synopsis?

Paul :p:
That's reason #2 to use a square wave. Reason #1 is to check for amplifier instability (and dirty pots and switches). Instability will show as ringing on the leading edge of the wave, and cruddy pots and switches will mangle the square wave so badly it will be unrecognisable.
 
You guys care to point me in the right direction regarding a sine/square wave generator?

Greg
 
Hey guys! Back. Yup, using square waves is a popular test. For frequency response, we tend to use sine waves. For linearity, we tend to use square waves. Root means fundamental frequency. Simply speaking, for an audio amplifier or other small signal amplifier, start at around 10Hz or so, or 20Hz or even DC if the manufacturer claims it and record both the input voltage and the output voltage unloaded and loaded sweeping to 20KHz and beyond to find the band pass (the flat frequency response) of the system and/or gain. Now that is neglecting a whole bunch of other qualifiers but hey, this isn't the lab. Now you can determine open voltage gain. In addition, with it loaded you can determine voltage gain loaded, current gain, and now power gain.

Now this may be a really old thread, but I have to chime in on this one for it has been a spot of contention between something mentioned in another thread and a repairman that unknowingly gave me my start. His name is Dave, and he has since passed; but I will always remember him. He worked at Hickok designing circuits which bear his name, a repairman in both the Jackson area and Greenwood. He gave me some wire wounds once for amplifier testing. I asked him about their inductance, remembering a comment here on AK about "not having to worry about some funky load you used". For the most part, those wire wounds do get HOT; all do at those power levels. It is just that, as mentioned, their resistance will vary with temperature. You will have to measure that temperature and the resistance after the test to see how close and to what degree you need to "pick off" the correct length for said power levels.

Now as far as wire anything goes, there will always be some inductance. Yup, I've got those shiny golden Litz Wound NH Dales; but at the end of the day, in the audio frequency band, as Dave pointed out, and I know now, it hasn't any bearing in repair work. It is usually negligible. Now say, you connect a bunch of them in parallel, the inductance drops even further. If they have the same inductance (same part hopefully close), the particular inductance at a frequency will half if two are placed in parallel and will be third if three...

Now about the above question: If you have a scope with two channels, on one channel measure voltage and on the other current. Measure the phase angle between them. That will give you your inductance with knowing the frequency of the drive. This is all basic EE stuff in Circuits II. I had a big one up on my mates and sometimes even the professors having dealt with Echowars and others like Dave when I started repairing scopes and stuff.

The information here is golden when it comes to this repair stuff. On a side not, I always used my ears for dirty pots. Now I have another method with the square wave drive. Nice!

Oh and for the record, I built a clone of the AudioGraph with some changes which was presented during my senior design evaluation/presentation. The IEEE guy said, "You know what he did, don't you? He automated an old school load box. He has the soul of an old engineer."

He spoke to me later about stuff way over my head afterwards in conference about complex impedance beyond just phase shifts, distortions, and frequency response. He said that they present all kind of nasties! One of these days...

Anyway, I still have the paper and the stuff; although, I did take it apart. It was heavy!!!!!
 
I have used Dale RH-250 - 8 Ohm, 1%, load resistors on my test bench for 40+ years for amplifier testing and evaluating with absolutely no problems. Worked as a service tech at a national distributor of well known and respected brands, and guess what was on my test bench as loads.....those same resistors! I mounted mine on a large heat sink to help them stay cooler. IMO it's the best way to go.
 
Back
Top Bottom