build a dummy load

[Randall]

I have a vintage Ampeg SVT I need to set the bias on. One of the 6550s white capped and flamed out one of the 3.6 ohm plate fuse resistors. It's back up and running, but I absent mindedly turned the bias adjustments when I was spraying them out (the amp was horridly grimy). It seems one needs a 350 watt 4 ohm dummy load to bias according to Ampeg instructions. The dummy loads I have seen are pretty spendy for how often I will need one.

SO, I'm thinking of buying 8 of these: THS502R0J TE Connectivity / Holsworthy | Mouser
2 ohm 50 watt resistors and wiring them up on a heat sink. One string of 4 in series = 8 ohms 200 watts. Wire another identical string in parallel with the first and I should have 4 ohms, 400 watts, correct?

Now my concern is, is this any more inherently risky then using one major wattage resistor? If any one of those 50 watters opens up, I'm down to 8 ohms 200 watts. Will I notice this if it happens? Because if I don't it won't take long to burn out the other side. My fear is suddenly going to no load under power. Bad news. My thinking is the 50 watters should not get overloaded because even if the amp is capable actually putting out 350 watts with used Russian 6550s, that is about 44 watts per resistor, or 88% of maximum rating.

Does this make sense?

[The Dude]

If you bias the amp by looking at idle current, you won't get anywhere close to needing a 400 watt load since this is done with no signal applied. You can use most any crappy speaker laying around that has the correct impedance.

[Randall]

So much for that idea...

[bob p]

Thinking outside of the load box

I've been having similar thoughts about a decent SVT load box. To test my SVTs at full drive into a dummy load, theory says that I'd need something that can handle at least 300W RMS at 4R or 2R.* (see below)

That does create a problem for me, as my loads are typically set up for 300W at 8R. I have a really nice 300W 10R vitreous enamel resistor that's tapped in 2R increments, so each successive tap adds 60W @ 2R. Although that's been more than adequate for most work, it falls really short of being adequate for the SVT.

I went ebay-ing and I bought a bunch of Ohmite vitreous enamel resistors. As it turns out 4R and 8R resistors are in demand, and tend to have high prices while 10R resistors are pretty cheap because nobody wants them. I ended up getting 10 of the Ohmite 10R / 100W vitreous enamel resistors for about $20. Putting 5 of them in parallel will give me a 500W / 2R load. Putting two of those arrays in series will give me a 1kW / 4R load. Add some switches or banana jacks and stuff it all into an old PC case and you've got the perfect SVT load box.

Or maybe you could just use a toaster.



* Now about theory vs. practicality. Vitreous enamel resistors are fragile, but if you mount them in a proper ventilated / screened case to protect them and to protect yourself from them, you can pretty much ignore their power ratings. Back in the day I saw guys in the radio R&D labs at Motorola driving vitreous enamel resistors so hard that they glowed bright orange. In other words, as long as you've got the right resistance rating, there's really not much limit to how much power you can put into it, assuming that you've got a safe installation that will keep you from getting burned and/or starting a fire and the resistance doesn't drift too much as the nichrome wire heats up. When maintaining constant resistance matters I've even seen guys overdrive the resistors in buckets of water and boil the water.

Along those lines, if I didn't have the 10-pack of Ohmite 270 series 10R resistors to make the big load box that I described earlier, I'd just plan on using a pair of properly mounted 100W 4R resistors to make a switchable 2/4/8-ohm load and not worry too much about how hot it gets when driven hard. Proper mounting is the key.

[Enzo]

A few thoughts...

The nice aluminum 50 watters do jump in price as you go over 50 watts. But instead of four 2 ohm in series twice, how about eight 8 ohm ones in series/parallel. Wire two in parallel, you now have a 4 ohm 100 watt resistor. Now three more such pairs, total four pairs, and wire them up series parallel as if they were a 4x12 speaker cab. Now if one were to open, you don;t lose half of them.

Plus when you want to spiff up the wiring to make it more versatile, I'd think 8 ohms might be more convenient. Use one for an amp head, or two, instead of a bunch of them. Make the investment a useful part of your shop.


But yes, your way is also correct, the math works.


I don;t think a single 400 watt resistor would be any more risky than 8 little ones. I don't worry about it at all. I am much more concerned that my wiring or the speaker cord or something like that will fail than I am the resistor. They are wirewound and sturdy. They take a lot, though I do have around here for show and tell one that actually melted. The aluminum flowed or softened and the thing slumped. AFter all, we play these amps into speaker cabs. Which do you feel more confident it won;t burn out, a speaker or an aluminum power resistor. If you are not afraid to connect a speaker cab to the amp, then a resistor bank ought not to scare you either. IN fact, you can pop a speaker with a spike a hell of a lot easier than one of those resistors.


Now, note that all the bias procedure does is measure cathode current through 1 ohm resistors, like a zillion bias probes. This is done at idle. The amp will not be producing 350 watts while you are doing this. They don't want you connecting anything to the amp that could not handle its full power, but full power is not involved in this procedure. presumably this amp has a speaker cab. And presume further the cab can handle the power from the amp. The amp cannot put out any more than its 330 watts, so if the cab can handle it, why not use the cab as the load. No parts ordering involved. Plug the speaker into the amp, let the amp warm up a bit to settle down, then adjust your bias on the little test points, and you are done. You won;t be playing the amp into the load to set the bias.

Of course if you want to do other tests without listening to a great noise, then the loads are great.


DOn;t get me wrong, having good dummy loads is a great step in building a shop, but if you are ordering them just for this project, well...

[The Dude]

Randall,

Please note: I edited/deleted part of the above post because of a severe error. Your combination should work. Although in a series bank of resistors the current is the same through the entire group, the voltage is divided among them. My apologies for the misstatement.

[Enzo]

Oh wow, between when I started writing and when I posted, three of you also posted, cool.

[The Dude]

Oh wow, between when I started writing and when I posted, three of you also posted, cool.

At least you didn't post something stupid.

[bob p]

we've all done it at least once.

[Gingertube]

I have 2 dummy loads

I was in the parts shop one day when I noticed a pack of 12 0R39 50W resistors in the specials bin. Bolted 10 of those on to a piece of C channel Aluminium Offcut (from the local aluminium suppliers offcut bin) and wired 10 of them in series. I have used it to test an SVT in continuous sinewave output power. It got a bit warm but that is all. THis was the MOSFET Output Unit and according to the Oscilloscope I got 431W RMS at just below clipping.

I have a decent hunk of heatsink, probably about 1 degree C per watt, to this I bolted 2 off 2R2 200W resistors in series. This is for testing lower power amps.

I need to knock up an 8 Ohm load. I guess at a pinch I could wire those 2 off 4 Ohm loads in series!!!!

As states above, if only setting bias currents then short the input and connect a simple resistor load, I generally use a 2W metal film for this. With no signal all you have is noise power. A 2 Watt is more than adequate (unless you really have a problem with the amp).

Cheers,
Ian

[R.G.]

Here you go, $0.02 a watt:

Two-ohm, 25W for $0.50.

Where you're trying to dissipate power, array solutions tend to be more reliable than point-solutions, as the failure of any one element doesn't change things much, and the temperature of any one element tends to be lower.

With a 2-ohm for the basic unit, you get four in series for 8 ohms/100W; two of those strings in parallel gives you 4 ohms/200W and four strings in parallel gives 2 ohms/400W for $8 for the resistors.

Identical-resistor matrices are favorite questions on sophomore EE mid terms.

[nickb]

Something I do nearly always is to use a pulsed test with 10% duty cycle. This means my 200W rated dummy can handle 2000W, or more usually just not run so hot. Snag is you do need a DSO but that's not unusual nowadays.

[bob p]

Here you go, $0.02 a watt:

Two-ohm, 25W for $0.50.

you can find some great deals on ebay. a while ago i bought a pile of the flat vitreous enamel stackable wirewound resistors in 15W, 15R. those ended up going into a 7.5R ladder attenuator. those 2R resistors look useful.

[bob p]

Something I do nearly always is to use a pulsed test with 10% duty cycle. This means my 200W rated dummy can handle 2000W, or more usually just not run so hot. Snag is you do need a DSO but that's not unusual nowadays.

i think that's a useful way to test your gear, but if used unscrupulously, a test like that can lead to the temptation to falsely inflate an amp's power rating.

i got kicked off of carversite because i complained that someone who was selling hot-rodded carver amps was being less than honest about what he was calling RMS power ratings on his hot-rodded amps. at one end of the scale, there's a guy like me who rates power using continuous sine wave drive, in the middle there's the FTC rating for RMS power, and at the other end of the scale there are guys who perform limited duty cycle, peak music power type testing, and pass off the numbers as if they were true FTC measurements of sustainable RMS power.

unfortunately, lots of hot-rodders like to compare apples to oranges. they like to compare their limited duty cycle, peak power test results to the RMS power ratings that were offered by the manufacturer, in order to give the illusion that their hot-rodded amps have more sustainable power than they really have.

unfortunately, most customers customers can't tell the difference in the ratings, and don't have the sophistication to realize that if the power supply isn't upgraded then the improved power ratings amount to smoke and mirrors. as time passes people also tend to forget all of the hoopla about fictitious power ratings in the 1970s that led the FTC to take action to regulate how amp makers specified their power ratings.

[Randall]

I realize setting bias on an SVT at idle according to the factory does not require a big wattage load, however many suggest adjusting crossover distortion at full power as the way to go. This is what I was concerned about. I think I will just set it with the cab plugged in with no signal. The phase inverter balance does call for being driven with a 40 Hz signal, however. I think I will connect a pair of my 8 ohm 300 watt bass cabs in parallel for that adjustment as well. I'll just have to pick the right time to do it. And make sure the toilet is unoccupied.

Thanks for the input, as always.