Moderate power amplifiers for continuous operation

[richardash1981]

I'm designing a test rig at work to test a piece of equipment designed to monitor the AC mains supply (240/415V). As such, we need a test AC voltage source where we can control voltage, frequency, waveform shape and so on, in a reproducible way. Given the last 12 months, we really want this to be a computer control, not a bunch of large handles turned by hand. This took me down the route of a PC, multi-output sound card, audio power amps and voltage step-up transformers (to get from lower voltage, high current, to higher voltage but less current). A prototype was duly produced with whatever was to hand (including some personal kit), and it works nicely, with each test case being a suitable sound file containing 50Hz-ish sine waves which gets "played" to the equipment under test. I'm now looking to buy a couple of power amplifiers to put this on a permanent basis, and probably a second rig.

The power amplifiers will be fed from the sound card line outputs (balanced TRS nominal +4dBu), and each of the three channels required drives a step-up transformer. The output voltage required from the amplifiers (three for three phase) is 18V RMS, feeding 18V:240V transformers. The loading is 15W worst case (it should be nearer 6W across three phases), so I make it that the amplifier is seeing an impedance of 22 ohms - higher than designed but not stupid. As a result, the amplifier needs to be rated at rather more than 15W into standard 8 ohms in order to get sufficient voltage swing. I make that a design power of 41W into 8 ohms to get the voltage swing.

 

The permanent rig will be in continuous operation - the idea will be to run in 24/7 for automated long term testing. And of course that means a continuous 50Hz tone going through the equipment, which is maybe not what it was designed for.

I originally looked at amplifiers for this before B**** made buying from Thomann.de messy, and was looking at two of their t.amp S-150 MK II amplifiers. I had looked at the 4-channel t.amp E4-130, but was put off by the idea of fans (not for noise, but because they will fill it up with dust, and wear out). Today (the job has come back) I went looking on plausible UK online retailers to see what they could do in the way of convection cooled amplifiers, and turned up the Fame Audio A-400 II on DV24/7. It looks suspiciously similar, and is rated at the same 150W into 4 ohms, ~75W into 8 ohms. I'm not that surprised that the same amplifier (undoubtedly imported from a contract manufacturer in the far east) shows up with multiple different brandings, but had started with Thomann because of good recommendations that the items they put their name on were generally good value. I've never heard of the Fame Audio branding before.

 

So two questions:

1. Does anyone have any experience of the DV24/7 "Fame Audio" amplifier, or what I presume to be their own-brand equipment in general?
2. Does the idea of running a convection cooled amplifier at half it's rated output voltage swing, and 10% of the notional output power, but on continuous tone signal sound plausible in terms of not overheating?

 

The other option at a similar price was a pair of Behringer A800. Also convection cooled, but with a switch mode power supply, which made me slightly wary about EMC issues.

[sunray]

I'm designing a test rig at work to test a piece of equipment designed to monitor the AC mains supply (240/415V). As such, we need a test AC voltage source where we can control voltage, frequency, waveform shape and so on, in a reproducible way. Given the last 12 months, we really want this to be a computer control, not a bunch of large handles turned by hand. This took me down the route of a PC, multi-output sound card, audio power amps and voltage step-up transformers (to get from lower voltage, high current, to higher voltage but less current). A prototype was duly produced with whatever was to hand (including some personal kit), and it works nicely, with each test case being a suitable sound file containing 50Hz-ish sine waves which gets "played" to the equipment under test. I'm now looking to buy a couple of power amplifiers to put this on a permanent basis, and probably a second rig.

The power amplifiers will be fed from the sound card line outputs (balanced TRS nominal +4dBu), and each of the three channels required drives a step-up transformer. The output voltage required from the amplifiers (three for three phase) is 18V RMS, feeding 18V:240V transformers. The loading is 15W worst case (it should be nearer 6W across three phases), so I make it that the amplifier is seeing an impedance of 22 ohms - higher than designed but not stupid. As a result, the amplifier needs to be rated at rather more than 15W into standard 8 ohms in order to get sufficient voltage swing. I make that a design power of 41W into 8 ohms to get the voltage swing.

 

The permanent rig will be in continuous operation - the idea will be to run in 24/7 for automated long term testing. And of course that means a continuous 50Hz tone going through the equipment, which is maybe not what it was designed for.

I originally looked at amplifiers for this before B**** made buying from Thomann.de messy, and was looking at two of their t.amp S-150 MK II amplifiers. I had looked at the 4-channel t.amp E4-130, but was put off by the idea of fans (not for noise, but because they will fill it up with dust, and wear out). Today (the job has come back) I went looking on plausible UK online retailers to see what they could do in the way of convection cooled amplifiers, and turned up the Fame Audio A-400 II on DV24/7. It looks suspiciously similar, and is rated at the same 150W into 4 ohms, ~75W into 8 ohms. I'm not that surprised that the same amplifier (undoubtedly imported from a contract manufacturer in the far east) shows up with multiple different brandings, but had started with Thomann because of good recommendations that the items they put their name on were generally good value. I've never heard of the Fame Audio branding before.

 

So two questions:

1. Does anyone have any experience of the DV24/7 "Fame Audio" amplifier, or what I presume to be their own-brand equipment in general?
2. Does the idea of running a convection cooled amplifier at half it's rated output voltage swing, and 10% of the notional output power, but on continuous tone signal sound plausible in terms of not overheating?

 

The other option at a similar price was a pair of Behringer A800. Also convection cooled, but with a switch mode power supply, which made me slightly wary about EMC issues.

Something like these: https://www.ebay.co....675.c101224.m-1

They are designed to run 24/7 and some have a 16ohm tap but please don't get hung up on impedance.

 

These come up regularly on ebay, look for 3 the same.

FWIW I use something similar for ringing a phone prop, albeit only for a few seconds

 

That seller has 2 at the moment.

 

Edited May 7, 2021 by sunray

[sandall]

Don't know that particular model, but the TOA amps I've had seem to be pretty bullet-proof.

[pmiller056]

I was involved with a similar, but larger setup many years ago with a rackful of Crown amplifiers driving 1kVA transformers. They amplifiers were badly damaged a few times because when the computer program was stopped, it froze the instantaneous output level into the amplifier. When this happened, the amplifier input effectively had a fixed DC input signal applied. Unfortunately the ampiifier was directly coupled through to the test rig transformer. A transformer will appear as a short circuit to DC. The amplifier overload protection to a DC fault was not great. When it did eventually cut off the output drive, the back EMF from the connected transformer destroyed the output stage. Once we understood what was going on, we were unable to significantly change the system due to time constrains, so we ended up making sure that all the drive signals to the amplifiers went smoothly to zero over several cycles when the control program was stopped or crashed.

In the light of my experience, I'd be inclined to go for some 100V line amplifiers because they are made to cope with transformer loads. Do check that the frequency response and available output power is adequate at your frequency of interest. You will need to get appropriate step up transformers to go from 100V to 240V rms. If you choose a conventional amplifier, add some flywheel diodes from the output terminal to the supply rails to protect the output transistors. In addiditon, put a suitable Varistor across the transformer primary to limit the magnitude of any back EMF under fault conditions.

 

[sunray]

I was involved with a similar, but larger setup many years ago with a rackful of Crown amplifiers driving 1kVA transformers. They amplifiers were badly damaged a few times because when the computer program was stopped, it froze the instantaneous output level into the amplifier. When this happened, the amplifier input effectively had a fixed DC input signal applied. Unfortunately the ampiifier was directly coupled through to the test rig transformer. A transformer will appear as a short circuit to DC. The amplifier overload protection to a DC fault was not great. When it did eventually cut off the output drive, the back EMF from the connected transformer destroyed the output stage. Once we understood what was going on, we were unable to significantly change the system due to time constrains, so we ended up making sure that all the drive signals to the amplifiers went smoothly to zero over several cycles when the control program was stopped or crashed.

In the light of my experience, I'd be inclined to go for some 100V line amplifiers because they are made to cope with transformer loads. Do check that the frequency response and available output power is adequate at your frequency of interest. You will need to get appropriate step up transformers to go from 100V to 240V rms. If you choose a conventional amplifier, add some flywheel diodes from the output terminal to the supply rails to protect the output transistors. In addiditon, put a suitable Varistor across the transformer primary to limit the magnitude of any back EMF under fault conditions.

 

It surprises me that happened with crown kit and equally there was not a single capacitor in the signal chain. One of the reasons I highlighted a 120W amp was to ensure the much lower power quoted by OP was not going to stress it and also the frequency would still within the frequency response curve.

 

Don't know that particular model, but the TOA amps I've had seem to be pretty bullet-proof.

The funny thing is I've always preferred Adastra but I have to say I only have 2 of them left out of possibly 12 over the years and they both have faults. The silly thing is they never seem right after the cover's been off.I have about 10 TOA and Inter-M and the only failures ever been a damaged pot (physically broken off) and a blown fuse when 24V applied reversed.

So yes I have to agree they are tough beasts but it doesn't stop my preferrence for the better input facilities, and especially levels, of the Adastras.

 

The sort of units I've linked to are all transformer output, the various output impedances are simply tapped off the output winding so it makes no difference, that I can see, where the additional transformer steps up from.

 

Edited May 8, 2021 by sunray

[Brian]

That seller has 2 at the moment.

 

 

If the OP goes that way then I have three 60W Millbank units sitting here doing nothing!

[sunray]

That seller has 2 at the moment.

 

 

If the OP goes that way then I have three 60W Millbank units sitting here doing nothing!

I'd have ripped your arm off a few years ago for those.

[richardash1981]

I was involved with a similar, but larger setup many years ago with a rackful of Crown amplifiers driving 1kVA transformers. They amplifiers were badly damaged a few times because when the computer program was stopped, it froze the instantaneous output level into the amplifier.

...

It surprises me that happened with crown kit and equally there was not a single capacitor in the signal chain.

The Crown DC-300 was of course DC coupled - they actually recommended an output DC blocking capacitor when using it bridged to direct-drive 70V line systems. This is one of the reasons I'm not looking to use the DC-300s which are in the store at work for this (as well as age and noisy/unstable gain controls).

 

Don't know that particular model, but the TOA amps I've had seem to be pretty bullet-proof.

The sort of units I've linked to are all transformer output, the various output impedances are simply tapped off the output winding so it makes no difference, that I can see, where the additional transformer steps up from.

That all makes sense. Church has been on the same TOA 100V mixer amp for longer than I have been around. The only problem has been an occasional open-circuit in the mains fuse holder which makes the whole device go dead! When you have the fuse out and back it then makes contact for another 5 years ...

 

[Simon Lewis]

A colleague worked on a research MRI scanner, with several Crown DC300s driving an improbably low resistance....

[richardash1981]

Looking at TOA specifications, their output transformers seem to be only specified down to 50Hz, and often -3dB there. This doesn't seem to bode especially well for working hard at 50Hz, which is certainly not what they are designed for.

 

One of the ranges seems to have a "Direct out" which bypasses the internal output transformer and significantly flattens and widens the frequency response:

https://www.toaelectronics.com/product/p-912mk2-120-watt-power-amplifierbut I'm not seeing this for sale in the UK and the https://www.toa.co.uk/products/amplifiers-1/power-amplifiers-analogue/ website doesn't list it. The similar P-1812 listed in the UK doesn't seem to have this feature, although the specification is rather vague - there is no amplitude tolerance given for the frequency response!

In some of the older TOA specs there are graphs. These show 50Hz response both rolling off and rising in distortion considerably - 2% THD at 1/3 of rated output ...

 

The step-up transformers we are using have a data sheet frequency range from 40Hz to 400Hz operation (torriods intended to work with mains and aircraft supplies) which is adequate for the purpose. I hadn't thought about the potential issue of back-emf from the transformer damaging the amplifier (I suppose I assumed that loudspeakers were equally unpleasant loads), so good point about over-voltage protection. You would hope that amplifier designers put flywheel diodes in as a matter of course, but I'm sure they don't.

 

A concern (and the reason I was looking at power amps rather than mixer amps on TOA website) is having a multitude of controls (including tone controls) in the mixer section in the signal path. As far as I can see TOA don't seem to have provided "mixer out / power amp in" or insert points on their design to make it possible to detach the power section from the mixer.

As this is work rather than me personally it is going to be easier to buy new (and there is some budget) than second hand unless there is a clear "you can't buy one new" case.

 

[sandall]

.... I have three 60W Millbank units

My all-time favourite for PA (& not just because I lived near Uckfield as a teenager)

[TomHoward]

For commercial there the InterM PA-9312 / 24 / 36 without the mixer section / tone controls

There’s also the QD-4480 / 4960 4-channel ones if you wanted to get it into one unit. The PA have transformers to 100V built in and the QD come with a matching 4-transformer bank so must surely be designed to drive them

 

Reading back apologies that those are all fan-cooled, but they are all commercial beasts. The PA certainly is and I am 99% sure the QD is as well

[pmiller056]

The Crown amplifiers I mentioned were actually Techron's. Techron specialises in big audio frequency power amplifiers for industrial and research applications. At the time we bought them (late 90's?), Techron was a subsidiary of Crown and fundamentally they were re-branded Crown Amplifiers. The LF frequency response may have been modified from the Crown design to go all the way to DC.

 

[sunray]

Looking at TOA specifications, their output transformers seem to be only specified down to 50Hz, and often -3dB there.

As mentioned earlier, that's the reason I suggested 120W units so you will be working way below the response curve. FWIW I've used the 1512 as a sub amp with a 100Hz crossover several times. Admittedly it doesn't compare with the likes of Eclar or Crown etc but it holds it's own

This doesn't seem to bode especially well for working hard at 50Hz, which is certainly not what they are designed for.

 

One of the ranges seems to have a "Direct out" which bypasses the internal output transformer and significantly flattens and widens the frequency response:

The mains/battery versions I been inside have push pull outputs working directly separately on 2 primary windings.

https://www.toaelect...er-amplifierbut I'm not seeing this for sale in the UK and the https://www.toa.co.u...fiers-analogue/ website doesn't list it. The similar P-1812 listed in the UK doesn't seem to have this feature, although the specification is rather vague - there is no amplitude tolerance given for the frequency response!

In some of the older TOA specs there are graphs. These show 50Hz response both rolling off and rising in distortion considerably - 2% THD at 1/3 of rated output ...

 

The step-up transformers we are using have a data sheet frequency range from 40Hz to 400Hz operation (torriods intended to work with mains and aircraft supplies) which is adequate for the purpose. I hadn't thought about the potential issue of back-emf from the transformer damaging the amplifier (I suppose I assumed that loudspeakers were equally unpleasant loads), so good point about over-voltage protection. You would hope that amplifier designers put flywheel diodes in as a matter of course, but I'm sure they don't.

 

A concern (and the reason I was looking at power amps rather than mixer amps on TOA website) is having a multitude of controls (including tone controls) in the mixer section in the signal path. As far as I can see TOA don't seem to have provided "mixer out / power amp in" or insert points on their design to make it possible to detach the power section from the mixer.

Just one of he features I like on some of the Adastra ranges!

As this is work rather than me personally it is going to be easier to buy new (and there is some budget) than second hand unless there is a clear "you can't buy one new" case.

 

How much engineering is possible? 100W class D amplifier boards are very cheap on Ebay now.

[jamesperrett]

I've been involved in using audio amplifiers for non audio applications and found that you need something pretty rugged if you want it to be reliable. We ended up using QSC RMX series amps which survived well when other lesser amps only lasted a few days. I would have thought that Crown amps should also be OK given that they are often used for vibrating tables and similar.