Speaker dummy load and effect ratings

[JHAnd]

Hi!

I want to build a dummy load box that I can attach various amplifiers. It will be used with both tube amplifiers and solid state amplifiers.

I'd like to know how to calculate what wattage my dummy resistors should be.

For example:

A 1973 Fender Twin tube amp is rated at 100W into 4 Ohms. Does that mean I need a 4Ohm 100W rated resistor as a dummy load? Or is there some safety margin factor that I should be aware of? Also would it be different with a 100W solid state amp or would I need the same 100W rated resistor? How long can a 100 W rated resistor be hooked up to a 100 W tube amp at maximum gain?

Also will two 8R 50W resistors wired in parallel be equal to a 4R 100W resistor?

Will two 4R 50W resistors in series be equal to an 8R 100W resistor?

In other words does resistor power rating add for both parallel and series connected resistors?


Thanks for any clarifying wisdom!

[Zero999]

The output power specification of an amplifier is often a lie. Quite often they'll specify the peak power for all channels, or just a fabricated number.

The maximum peak power output is simply: P = V_OUT²/R_L. V_OUT is normally a few volts under the internal power supply to the output transistors. With glassware, the speaker is driven via a transformer, so the output voltage is supply voltage, minus the losses, divided by the turns ratio of the transformer.

In reality, the speaker doesn't dissipate the peak power all of the time. The average power, before clipping which causes distortion, even at full volume, is tiny fraction of the peak power.

[TimFox]

It also depends on why you use a dummy load.
If you are trying to measure full constant power, or the increase in THD with increasing output, the resistor may need to handle the full power, even if the rated power is "optimistic".
For guitar amplifiers, which can be driven past high-distortion into serious clipping, the true mean power will be dissipated in the resistor for a steady-state case.
However, practical guitar and hifi amplifiers often can put out full power for only a very short period due to finite-size filter capacitors in the power supply.

Note that many high power resistors in metal cases with screw mount are rated like power transistors:  the "100 W" rating may imply a reasonable case temperature, which is determined by the heat sink.
For a Vishay/Sfernice RH50 resistor, it is rated for 50 W up to 25 deg C ambient temperature, but on a heat sink 50 W will raise the "hot spot" temperature to 150 deg C, and the resistor will die at 200 deg C.
That data sheet hard to parse:  for the RH50 it also shows an unmounted power rating of 9.6 W at 75 deg C ambient and 40 W on a 536 cm² area heat sink at 75 deg C ambient.
https://www.vishay.com/docs/50013/rh.pdf

The Vishay/Dale/Huntington AH100 is rated for 100 W on a "standard heatsink" (1000 cm² of 3 mm thickness), but only 50 W unmounted.
https://www.vishay.com/docs/31883/ah.pdf

[langwadt]

The output power specification of an amplifier is often a lie. Quite often they'll specify the peak power for all channels, or just a fabricated number.

The maximum peak power output is simply: P = V_OUT²/R_L. V_OUT is normally a few volts under the internal power supply to the output transistors. With glassware, the speaker is driven via a transformer, so the output voltage is supply voltage, minus the losses, divided by the turns ratio of the transformer.

In reality, the speaker doesn't dissipate the peak power all of the time. The average power, before clipping which causes distortion, even at full volume, is tiny fraction of the peak power.

afaik the crest factor for most music is in the order of 10-20dB, a sinewave 3dB

so an amplifier with the powersupply, heatsink etc. to do a continuous full power sinewave would be overkill in most cases

[Benta]

You're in Sweden, where electrical heating is pretty common.
I'd simply take a 1200 W heater panel from a dumpster somewhere and use it as load.
Heaters are (unlike incandescent bulbs) pretty linear with temperature.
That should take care of your 4 ohm load. For 8 ohms, use two 1200 W heaters in series.

If you want exact readings, do that with your instruments.

[TimFox]

The main use I have had for high-power dummy loads is testing amplifiers for THD vs. output power, using a continuous tone input.

[alligatorblues]

Resistors in series give 4R 50W +4R 50W = 8R 50W, in  parallel 2R 100W. The gold aluminum housed resistors with high power ratings usually have a flat side with 2 holes  in it for mounting on a heat sink. You can also add a fan. I use 300W Ceramic type made by Ohmite. Those just need some  nuts and bolts for the terminals #8, and an Ohmite bracket. They cost about $30, or you might get lucky on eBay.  These need heat sinks:
https://www.ebay.com/itm/125533177892?hash=item1d3a5c4824:g:fmEAAOSw9XxjMmEd&amdata=enc%3AAQAIAAAA4AWQu7Nj4wdVE%2Fd96tRqIeyEcypLgiAg8452ggINsmfH3Wt8qJfw1%2FpTwTg7OegM6CQe9XWeLf9y2jCNRUil1X3Z%2FZ%2B3ew1Gv8bjCiYxl5zGluhfphrDrqHqejpbImwftcmosRwkG6i%2F4dB3NdQm1DGff%2Boka2npusrH2L30I224sfsuZ6AcYsRPpU8E7Ibhom25Ea40Nsh68GkPz7ZtFYtOvCCyBxkNFcAxWzw86VmVsCgiqyrEXxMjYLEZ7ZZN8nSQ3n6sF9Prtq%2Frct25vCwGws81OZIWAUbjbaHwINVa%7Ctkp%3ABFBM-Mmb4OZi

These dont: https://www.ebay.com/itm/364213623778?hash=item54ccd2dbe2:g:y7AAAOSwEONkM7Np&amdata=enc%3AAQAIAAAA8Pu4GaL0MvTguQdNciKrTYdAcP4FhBNgODF1PSIvUqMstKJhxZkJPArV6JHIx4QNsNfzg5d6FWyQTLNFoUtcrLMbfnfuTNa5JoCBLC0MvxmeqOkmIjA%2FU4sflCl69ZhUI78BInnq2i%2BB3YS90HAahVNJ9mvw0FIjX9%2Fpi8U1vHRL6yaKxdRHL392AMvWKAKfgX6PtdBiDlMByQ9eERPfYz%2FMtmGlCG9HtE8ktLLbNkZPFbBSRPImvcSMMn9Yf15dOjATQ4ScA7%2BXRiOB8qk9dibgZB73Z5Brg7IMzWR%2B8Vt%2BPGxnsfMZjjUoCjde0yGNgw%3D%3D%7Ctkp%3ABFBM_o294OZi

[Swake]

Resistors in series give 4R 50W +4R 50W = 8R 50W.

Each resistor can still dissipate 50W, this becomes 8R 100W.

[langwadt]

Resistors in series give 4R 50W +4R 50W = 8R 50W, in  parallel 2R 100W. The gold aluminum housed resistors with high power ratings usually have a flat side with 2 holes  in it for mounting on a heat sink. You can also add a fan. I use 300W Ceramic type made by Ohmite. Those just need some  nuts and bolts for the terminals #8, and an Ohmite bracket. They cost about $30, or you might get lucky on eBay.  These need heat sinks:
https://www.ebay.com/itm/125533177892?hash=item1d3a5c4824:g:fmEAAOSw9XxjMmEd&amdata=enc%3AAQAIAAAA4AWQu7Nj4wdVE%2Fd96tRqIeyEcypLgiAg8452ggINsmfH3Wt8qJfw1%2FpTwTg7OegM6CQe9XWeLf9y2jCNRUil1X3Z%2FZ%2B3ew1Gv8bjCiYxl5zGluhfphrDrqHqejpbImwftcmosRwkG6i%2F4dB3NdQm1DGff%2Boka2npusrH2L30I224sfsuZ6AcYsRPpU8E7Ibhom25Ea40Nsh68GkPz7ZtFYtOvCCyBxkNFcAxWzw86VmVsCgiqyrEXxMjYLEZ7ZZN8nSQ3n6sF9Prtq%2Frct25vCwGws81OZIWAUbjbaHwINVa%7Ctkp%3ABFBM-Mmb4OZi

These dont: https://www.ebay.com/itm/364213623778?hash=item54ccd2dbe2:g:y7AAAOSwEONkM7Np&amdata=enc%3AAQAIAAAA8Pu4GaL0MvTguQdNciKrTYdAcP4FhBNgODF1PSIvUqMstKJhxZkJPArV6JHIx4QNsNfzg5d6FWyQTLNFoUtcrLMbfnfuTNa5JoCBLC0MvxmeqOkmIjA%2FU4sflCl69ZhUI78BInnq2i%2BB3YS90HAahVNJ9mvw0FIjX9%2Fpi8U1vHRL6yaKxdRHL392AMvWKAKfgX6PtdBiDlMByQ9eERPfYz%2FMtmGlCG9HtE8ktLLbNkZPFbBSRPImvcSMMn9Yf15dOjATQ4ScA7%2BXRiOB8qk9dibgZB73Z5Brg7IMzWR%2B8Vt%2BPGxnsfMZjjUoCjde0yGNgw%3D%3D%7Ctkp%3ABFBM_o294OZi

make sure to read the datasheet, for derating when it is not 25'C ..

[WimWalther]

Here's how I made my audio dummy load. A pair of salvage heatsinks each with 2X 4R & 2X 8R 100W 1% resistors. Note that these are CHINESE "100W" parts, so they're a fair bit smaller (and much cheaper) than their Dale / Vishay cousins. But in intermittent use, on decent sinks, they serve just fine.

They're set up like this:

4R.  4R
8R.  8R

4R   4R
8R.  8R

With this parts complement, you can configure values of 1, 2, 3, 4, 6, 8, 10, 12, 16 or 32R always of at least 100W (often much more) by changing jumpers around. I already had the sinks, and the resistors were about $2/ea. Cheap & cheery!

[CaptDon]

I think you need to spend an entire evening drawing some resistor circuits on paper both series and parallel and in your calculations simply use a fixed D.C. voltage, perhaps 16 volts and do your ohms law calculations until you understand how power adds or divides. Ohms law is simple and basic and if you don't have a full grasp of it maybe you better stop for a minute before you wreck something valuable. Also, in my lab I use the big ceramic resistors guessing about 10 inches long and I am lucky to have the ones marked N.I. or 'non-inductive'. You can spot an N.I. wirewound tubular resistor by the way the wire is wound back on itself instead of just one simple winding end to end. Speakers are highly inductive and a resistor with self inductance isn't harmful and may even better simulate an actual speaker load but I prefer the N.I. style for laboratory quality measurements and certification proving that repairs 'shall meet' published specifications. Some of my loads are fan cooled, some are in an oil bath. I routinely work on a particular model of amplifier that does 80vrms sinewave across 4 ohms non-inductive at 1600 real watts of heat in the load. That same amplifier can do bursts of tone at 100vrms across 4 ohms for a burst output of 2500 watts. To meet spec it must be able to do both.

[donlisms]

Indeed - plenty of good advice.  I think the most important bit is to realize those aluminum-housed resistors assume a good heat sink for the rating given.  Without, the rating is significantly lower.  With a fan, you're in m h better shape.  With cheap resistors from the bay, their ratings are probably optimistic even with a heat sink.

So... if you want a nice 100 ohm load, get 300 watts worth of resistors, put them on a heat sink, put a fan on it, and do not worry too much.

[WimWalther]

Don't worry about non-inductive resistors for your uses. Typical 4-8R parts don't have much inductance to begin with, and the effect at audio test frequencies (400Hz, 1KHz, etc) is negligible.

[wasedadoc]

You're in Sweden, where electrical heating is pretty common.
I'd simply take a 1200 W heater panel from a dumpster somewhere and use it as load.
Heaters are (unlike incandescent bulbs) pretty linear with temperature.
That should take care of your 4 ohm load. For 8 ohms, use two 1200 W heaters in series.

If you want exact readings, do that with your instruments.

That is all wrong. 230 Volts and 1200 Watts is 5.2 Amps and 44 Ohms.

[coromonadalix]

i use 4 ohms 200w aluminum sinked resistors and put them in a paint can with oil to help dissipate  loll

[JHAnd]

Thank you for all the info, I really appreciate all of your answers!

I was thinking of using four 4R "ARCOL HS100" 100W resistors.
https://www.elfa.se/sv/tradlindat-motstand-100w-4ohm-arcol-hs100-4r/p/30115980

Then I can get these configurations:

4R 400W
8R 200W
16R 400W

ARCOL Datasheet: https://media.distrelec.com/Web/Downloads/_t/ds/HS%2012-14.08_eng_tds.pdf

According to the Datasheet, If I'm not mistaken, I would need one 995cm2 3mm heatsink / 100W resistor.

However they seem to sell a heatsink that is made for the HS100
https://www.elfa.se/sv/kylelement-till-arcol-motstand-hs100-3w-ohmite-ah50600v05000fe/p/30198569

Four of these, one for every resistor would amount to 155 Euro. Plus 38 Euro for the resistors would be 193 Eur.

That seems a bit too much to me. Perhaps I could cram two resistors into one heatsink and hope for the best.

Is it really that important with the heatsinks?

I watched a video with D-Lab Electronics. He used four 4R 50W resistors and mounted them on a piece of sheet metal.
Link to video:

Perhaps if I need to use the maximum volume I could do it for shorter periods test quickly and then lower volume.

Would that be a viable strategy?

A typical amp that I would test would be the fender twin reverb 100W tube amp that wants a 4Ohm load.
also a 400W solid state bass amp that wants an 8Ohm load.

I might seldom need to test the amps full volume. Perhaps up to clipping point. I have a 40W tube amp right now that I'm going to have to test it's full range though. And I don't know what the future holds.

[TimFox]

Short tests:  that depends on what measurement you are doing for the test.
On a single-sweep DSO measurement, you only need a short "key down" time, but with a conventional THD meter you may need a longer time.
To be safe, you might want a thermometer handy, rather than wait for the "smoke test" to complete.

[alligatorblues]

Resistors in series give 4R 50W +4R 50W = 8R 50W.

Each resistor can still dissipate 50W, this becomes 8R 100W.

That was my idiot move of the day!

[Vovk_Z]

Four of these, one for every resistor would amount to 155 Euro. Plus 38 Euro for the resistors would be 193 Eur.
That seems a bit too much to me. Perhaps I could cram two resistors into one heatsink and hope for the best.

It seems to me too expensive too. Possibly they are expensive because those resistors have quite low TCR (< 100 ppm < 1k)? (IDK, I haven't used them).
TCR may became important if you want to measure ultra-low THD (near 0.001%).
If you are going to measure ultra-low THD (0.001%) then you need oversized resistors so they weren't too hot with a peak power (you want them to remain as linear as possible == as cold as possible).

[TimFox]

With a typical feedback-controlled amplifier that has a low output impedance, the small changes in resistance due to temperature variation during the periodic signal will have less than proportional effect on the THD seen in the voltage across it:  an interesting quantitative computation question.

[Vovk_Z]

An amplifier itself may have quite low output impedance but it will typically have some additional wires out of feedback loop. E.g. microhenry output filter or some protective relay circuit, and don't forget commutating wires, especialy wires to the dummy load itself. It all adds to quite non-zero value so it can hudgely worsen the THD if ampifier is zero-THD itself.

[TimFox]

Obviously, if the amplifier has zero distortion, anything non-linear after it will increase the distortion dramatically over zero:  "divide by zero error".
My point is that for a practical amplifier, with feedback that reduces its output impedance as seen at the output terminals to a very low value, a small variation of the temperature of the load resistor's instantaneous resistance during the signal period will have less than proportional effect on the harmonic content of the voltage across the resistance.
Besides the voltage-division effect of the low (but finite) output impedance and the low (but finite) variation in the load, there is also the thermal effect that the temperature excursions of the winding wire (or other element) thermally coupled to a heat sink (or even just the case and core) will be reduced by the thermal capacity of the materials.
Therefore, if the resistor is characterized at DC by, say, 100 ppm/K variation, that should typically induce much less than 100 ppm THD in a practical situation.
Two 12-inch (30 cm) AWG18 (roughly 1 mm diameter) wires from the amplifier to resistor is 13x10^-3
100 ppm of 8\$\Omega\$ = 0.8x10^-3
However, the voltage division ratio in this quantitative example is 8/8.013 = 1 - 1.6x10^-3
The rest is left as an exercise for the reader.

[BlackICE]

Electrodes in bucket of saltwater. Do it outside!


https://www.diyaudio.com/community/threads/power-load-dummy-load-pic.106261/

Two plates immersed in salt water can be used as a dummy load for high power.

Indeed, if you have use for the chlorine that gasses out. Your lungs and most anything organic around you (the dreaded "enviroment") don't.

nevertheless, mains connected saltwater tubes were used as heaters once

mains connected saltwater tubes were used as heaters once
X-rays were used to fit shoes once.
Asbestos was used in brake discs once.
Mercury was used to "cure" syphilis once.

massdevelopment of chlorine is limited to DC, in case the saltwater is used as AC-load only, there may be small amounts of chlorine escaping the water
if this leads to a dangerous level in a well vented enviroment, is at question.
I suppose saltwater is a bad dummy load for amplifiers, because of the large! tempco, the frequency behaviour and whatever.
regards

[langwadt]

https://youtube.com/shorts/DTuyDAh7NMM

[JHAnd]

So... If I don't want to spend money on four heatsinks for four 100W resistors...

Can I use a small bucket of cold tap water as a heatsink and  just lower the dummy load resistors into the water?

I'm not talking salt water, like the posts above... just regular tap water.