AF Power meter

[govjohn]

Hi, I wonder If anyone can explain exactly on what or why a piece of test equipment I have come across would be used? It is a MARCONI INSTRUMENTS AF Power meter 893B. It has limited power handling ability (10W) and on its own cannot indicate the 'quality' of the signal its measuring. It would have to be used with a 'scope or distortion test set and then you may as well use a dummy load which would be a hell of a lot cheaper. Look at its construction, it contains more precision resistors than I have seen together in my life! It must have cost a fortune especially being made by MI. (calm down John) Looking at its labels it was used on a government project, would these have been used by anyone else if so why? The dates on my camera appear wrong.
John.

[Station X]

I use one to assist in aligning radios and to check the output power of amplifiers. Some versions were fitted with a SINAD filter to aid FM alignment.

[Radio Wrangler]

It is used to do checks of receiver performance against specifications.

Can the receiver output the required audio power to speakers and/or headphones?
Is that audio power flat enough over the required audio band as the modulation source of a sig gen is swept?
Measure the noise level, then apply a signal with specified modulation and find out how much signal is needed to get the audio a specified number of dB over the noise level.
Then if it has the SINAD filter, put the sig gen to the appropriate modulation frequency, then by flipping the SINAD filter in and out measure the signat-to-(noise and distortion) ratio. Find out how much RF power from the sig gen is needed to get a specified SINAD value.

SINAD is specified on almost all professional radios, AM, FM, SSB...

Tedious work, now much automated. I use an HP98903B, now a classic itself, but it measures audio levels, sig/noise ratio, Sinad and distortion at the same time.

THese tests get done to qualify new equipment designs, to qualify new units at the end of the production line, in repair shops when someone's taken a transceiver out of a plane/tank/whatever and sent it back in for repair saying things couldn't be heard.

David

[govjohn]

Thanks for speed of the reply. My question concerns the expence of purchasing one of these where any facility capable of using for it the purposes mentioned would surely have better equipment avilable already as in the case of a receiver, it is not just the quantity of the output but the quality that is important and this instrument (nice though it is ) can give no indication of it.

John.

[Station X]

We're talking radio receiver alignment here, not Hi-Fi. Horses for courses.

AF output meters have been around for years. A homebrew version is described here:-

http://www.vintage-radio.com/project...indicator.html

[G0HZU_JMR]

Originally Posted by govjohn

Thanks for speed of the reply. My question concerns the expence of purchasing one of these where any facility capable of using for it the purposes mentioned would surely have better equipment avilable already as in the case of a receiver, it is not just the quantity of the output but the quality that is important and this instrument (nice though it is ) can give no indication of it.

John.

I think the value of the instrument is in its versatility because it supports a huge range of test impedances. On a service bench it would allow the operator to quickly select the relevant impedance in ohms and then get a readout on the dial for AF power into that impedance. It can also support a balanced input.

Apart from that, there isn't much to like because the detector is probably just a diode bridge (not true Vrms) and the datasheet shows that the optional SINAD filter has poor specifications in terms of notch depth and width.

This will all add a fair bit of measurement uncertainty but this would still be OK if someone was testing for basic compliance and a suitable pass margin was included to offset the measurement uncertainty.

Despite the classic 1980s Marconi enclosure I think the innards are based on a very old design.

[Dickie]

Yep, based on the TF893 from 1959, if not earlier. The SINAD filter was added much later in the 80s to prop up its sales. The meter is entirely passive and the filter is a simple twin T notch.

[David Simpson]

Marconi's range of small AF absorption wattmeters over the years(CT44 - who remembers them ?) have been jolly robust & reliable. In the RAF, in the 60's, we used the CT44 for testing intercom amps & transceiver audio amps, etc. Obviously they are no use for powerful stereo amps or huge big guitar amps. But ideal for vintage domestic radios. Yours has an excellent range of impedances. I.e. To get a correct measurement you need to chose an impedance which matches the o/p impedance of the audio circuitry being tested. The secondary of a speaker transformer for example.
A nice addition John for your workshop.

Regards, David

[G0HZU_JMR]

Looking at the dial scaling on the 893B I would guess that the (bridge?) diode detector is being run at a very low level close to the square law region. The datasheet does mention that the detector is temperature compensated and this compensation would definitely be needed when running at such a low detector level.

The dial scaling for power isn't linear but it isn't that far off and by running the detector in the transition region (just above square law?) it would give the meter some immunity against uncertainty due to any harmonic energy.

This would mean it could also measure the combined power of noise and harmonics with much less uncertainty when compared to a classic high level diode detector.

However, I can only guess at the above based on what it says on the datasheet and on the internal and external images posted earlier.

[Dickie]

This should answer all of your questions!

[Diabolical Artificer]

Could it be used too to find the OP Z of AF stages, by selecting Z till the OP drops by half?

Andy.

[Dickie]

More likely to vary Z until o/p power reaches a maximum?

[Herald1360]

Originally Posted by Diabolical Artificer

Could it be used too to find the OP Z of AF stages, by selecting Z till the OP drops by half?

Andy.

If you dropped the load R that low, for most tranny amps you'd be letting out the magic smoke!

Usual confusion between OP Z and design load Z?

[G0HZU_JMR]

Originally Posted by Dickie

This should answer all of your questions!

Thanks! That shows a back to back (1N4148) diode based protection clamp D7 + D8 ahead of the D1 to D4 detector and the detector drives the meter directly. This shows that the detector must operate at a low AC voltage drive level. The detector diodes are HP 5082-2835 Schottky diodes and these must be operating just above the square law region in this circuit.
The text in the diagram below states that the detector responds to the average voltage.

I made up a crude SPLINE model based on the HP 5082-2835 datasheet curve and the diode equation and I ended up basing the SPLINE data on a curve set by N=1.1 and Is = 2.2x10e-8 A and this agrees with the datasheet figures. Putting this into a model of the 893 detector circuit showed a detector response that was reasonably close to square law and with similar scaling to the meter dial.

It's impressive that they can get this to behave over temperature with the 15k temperature compensation resistor.