I'm building a quad ARL (Aiken Reactive Load) speaker emulator / dummy load / power soak. It's 4 resistive dummy loads and 4 sets of 'reactive' components, with banana patch board for all components, and some 1/4" jacks to banana jacks, with big heat-sink, big inductors... each somewhat like this but voiced to emulate a specific speaker/cab combination and built for 8 ohms:
http://i557.photobucket.com/albums/ss18/ilya-v/Aiken.jpg
Instead of a nonpolar electrolytic or two electrolytics back-to-back, I bought 4 poly/metal-film motor start capacitors: AC Motor Capacitor Start Capacitor CD60 450VAC 200uF | eBay
They're 200uF and 450 VAC. My concern is that they're motor-START capacitors, and not motor-RUN capacitors. The start caps are meant for only occasinal intermittent duty, typically less than 2 seconds. Run continuously they overheat, smoke, leak, burn...etc. In this application there is a resistor and inductor in parallel with this cap, and that network is in series with another and with the resistive load. The 450v. rating seems like I shouldn't have to worry about using it even with the high power content of a square wave from a clipping guitar amp???? Yes? No? I don't have Spice or the inclination to do the math...but I wouldn't expect anywhere near that. Am I OK?
http://i557.photobucket.com/albums/ss18/ilya-v/Aiken.jpg
Instead of a nonpolar electrolytic or two electrolytics back-to-back, I bought 4 poly/metal-film motor start capacitors: AC Motor Capacitor Start Capacitor CD60 450VAC 200uF | eBay
They're 200uF and 450 VAC. My concern is that they're motor-START capacitors, and not motor-RUN capacitors. The start caps are meant for only occasinal intermittent duty, typically less than 2 seconds. Run continuously they overheat, smoke, leak, burn...etc. In this application there is a resistor and inductor in parallel with this cap, and that network is in series with another and with the resistive load. The 450v. rating seems like I shouldn't have to worry about using it even with the high power content of a square wave from a clipping guitar amp???? Yes? No? I don't have Spice or the inclination to do the math...but I wouldn't expect anywhere near that. Am I OK?
Last edited:
You got the cheapest quality of motor start capacitor, and this eventually will be a problem.
Many industrial electricians including me, we do straggle to get NOS old/new stock, made in USA or EUROPE capacitors, because the modern ones in China has a tremendously low survival rate.
Many industrial electricians including me, we do straggle to get NOS old/new stock, made in USA or EUROPE capacitors, because the modern ones in China has a tremendously low survival rate.
Agree in general, but I guess that any amplifier you throw at it will be nothing compared to the stress of starting a motor, which runs on 220V, is purely inductive, has stall current much higher than run current, kicks back like a mule, etc.
Compared to that, audio load service is light.
Voltage wise how much will a 200W/8r amp send? ... 40VAC?
Only potential problem is that it's part of an oscillating tank and that will both raise voltage and current, but anyway tank Q is limited, both by the 8 ohms series resistor and mainly the 33 ohms parallel one.
To be certain, build one, connect an SS amp to it, say one capable of 20V RMS output (50W/8r) and sweep it from 200Hz down.
You will find a peak voltage across the cap at resonance.
How much does it rise? .... 30V .... 40V? .... 50V?
Whatever you measure now you know the resonance increase factor, call it "K" or whatever , and you'll know that at resonance, with 40VAC input expected , it will rise at most to K*40VAC.
Very much doubt it reaches dangerous levels and to boot, you will be playing a guitar, not feeding continuous signal at resonance, so you will have an additional safety factor.
Please post your measurements here, they will be very useful to others.
Very much liked the Aiken load 🙂
Compared to that, audio load service is light.
Voltage wise how much will a 200W/8r amp send? ... 40VAC?
Only potential problem is that it's part of an oscillating tank and that will both raise voltage and current, but anyway tank Q is limited, both by the 8 ohms series resistor and mainly the 33 ohms parallel one.
To be certain, build one, connect an SS amp to it, say one capable of 20V RMS output (50W/8r) and sweep it from 200Hz down.
You will find a peak voltage across the cap at resonance.
How much does it rise? .... 30V .... 40V? .... 50V?
Whatever you measure now you know the resonance increase factor, call it "K" or whatever , and you'll know that at resonance, with 40VAC input expected , it will rise at most to K*40VAC.
Very much doubt it reaches dangerous levels and to boot, you will be playing a guitar, not feeding continuous signal at resonance, so you will have an additional safety factor.
Please post your measurements here, they will be very useful to others.
Very much liked the Aiken load 🙂
If you hear a hissing sound (whistling) sound from the amp -not the speakers TURN OFF THE AMP. The capacitors are about to blow.
I work in hvac and those things can go violently!
I work in hvac and those things can go violently!
If you could guess the true service life in hours of that thing? I would pay you in cash.
You made a soup of thoughts by blending terms as oscillating and resonance, when there is about six true electrical parameters which are measurable by an LCR meter, and they can offer true evidence of what is happening inside to any capacitor.
I am not saying all this as been a teacher talking to student, because I am not teaching anything due internet.
But when some one expressing fear about the performance of his project he expects in return more facts as answers than speculations.
Yes, there are youtube videos of how the caps blow if you put 60hz on them continuously instead of intermittently, with a motor-start cap put into motor-run duty. So I got the 450 volt version instead of the 250 volt version.
If I just try a destructive test the big special-core inductors will saturate, but a stress test would tell me a lot. Fahey's got a good idea, but I still would need to blow up a cap in order to find out what it can take, as its intermittent-duty ratings have little relevance. My cheapo meter, measuring at 600Khz gives:
192 uF 4% low
212 uF 6% high
209 uF 4.5% high
195 uF 2.5% low
So that's acceptable, and better than motor-run caps which tend to vary even more and are often rated with a range of values the cap can replace on a motor, rather than any specific number. But at this price I could have bought several more and put the outlying ones away for some other project.
I'm just going to build it, and if it blows I'll get motor-run caps next time.
I should do Fahey's test in a safe power range, and extrapolate that to a power range where the inductors start to saturate, then do an additional destructive test and see where one actually blows up.
Or, I can just build a complete circuit on the bench and drive it overnight with a square wave at the high end of the range I intend to use it in.
There are also inexpensive motor-run caps available, and next time that's what I'll get.
If I just try a destructive test the big special-core inductors will saturate, but a stress test would tell me a lot. Fahey's got a good idea, but I still would need to blow up a cap in order to find out what it can take, as its intermittent-duty ratings have little relevance. My cheapo meter, measuring at 600Khz gives:
192 uF 4% low
212 uF 6% high
209 uF 4.5% high
195 uF 2.5% low
So that's acceptable, and better than motor-run caps which tend to vary even more and are often rated with a range of values the cap can replace on a motor, rather than any specific number. But at this price I could have bought several more and put the outlying ones away for some other project.
I'm just going to build it, and if it blows I'll get motor-run caps next time.
I should do Fahey's test in a safe power range, and extrapolate that to a power range where the inductors start to saturate, then do an additional destructive test and see where one actually blows up.
Or, I can just build a complete circuit on the bench and drive it overnight with a square wave at the high end of the range I intend to use it in.
There are also inexpensive motor-run caps available, and next time that's what I'll get.
Oh come on, Fahey gave the most practical conservative advice on how to measure it in the circuit in a safe range then extrapolate it to where I will run it. WHo can know the service life there? Only service life testing, and I'm not inclined to wait that long. I don't see kiriakos drawing the curves of voltage and current across and thru the cap and I still would not understand the phase relationships clearly. I need to learn to use spice, it would make my life easier. But Fahey has been incredibly helpful to me, and I'm just grateful to listen and learn when he posts.
The idea of having the components for four ARL sections on a banana-plug patch board is that I can use them wired up on the patch bay just like drivers in a cab, and make various impedance loads and I can tap off to drive various impedance real speakers, adjustable resistive loads like a power soak or some high-power variable resistors or L-Pads.
When I have the patch bay wired like a 4X12 bottom, I can remove one resistive section and in its place patch in a 1/4" jack to cable to any commercial adjustable power soak. Each ARL will be only lightly loaded, and so will the outboard power soak.
The idea of having the components for four ARL sections on a banana-plug patch board is that I can use them wired up on the patch bay just like drivers in a cab, and make various impedance loads and I can tap off to drive various impedance real speakers, adjustable resistive loads like a power soak or some high-power variable resistors or L-Pads.
When I have the patch bay wired like a 4X12 bottom, I can remove one resistive section and in its place patch in a 1/4" jack to cable to any commercial adjustable power soak. Each ARL will be only lightly loaded, and so will the outboard power soak.
Last edited:
Heck, the way I'm going to run it, it is so incredibly overbuilt... When I'm using all 4 reactive sections, combined that network will probably handle an SVT or a Trace Elliot V8 in full clip. There's really only one way to find out. I'm hoping it might be possible to feel them get warm before any damage occurs.
Just ball-park numbers I'm figuring a more typical 100-watt+ amp making square waves, maybe 40 volts across it at the worst frequency if I used just one section, or 20 volts and half the current on each if using all 4 sections wired parallel/series...with caps rated at 450 volts. That's 22 times the expected voltage. And even if the amp is making square waves with droning death-metal (worst case LOL) and the drive level really is continuous, the frequency content will be shifting around (assuming it's not just one note LOL) in & out of circuit resonance and not locked at the worst frequency the caps absorbs. This is safe enough for me to construct and test and measure.
Thanks everyone.
Just ball-park numbers I'm figuring a more typical 100-watt+ amp making square waves, maybe 40 volts across it at the worst frequency if I used just one section, or 20 volts and half the current on each if using all 4 sections wired parallel/series...with caps rated at 450 volts. That's 22 times the expected voltage. And even if the amp is making square waves with droning death-metal (worst case LOL) and the drive level really is continuous, the frequency content will be shifting around (assuming it's not just one note LOL) in & out of circuit resonance and not locked at the worst frequency the caps absorbs. This is safe enough for me to construct and test and measure.
Thanks everyone.
Last edited:
- Status
- Not open for further replies.