PEP? Are Those Power Meters Showing Us PEP Peak Envelope Power, or just Peak Power?

Discussion in 'Amateur Radio Amplifiers' started by KQ4X, Dec 31, 2012.

Thread Status:
Not open for further replies.
ad: L-HROutlet
ad: l-rl
ad: Left-3
ad: Left-2
ad: Subscribe
ad: L-MFJ
ad: MessiPaoloni-1
  1. KB5UBI

    KB5UBI Ham Member QRZ Page

    I can't comment on the MFJ-989C, but the MFJ-989D does indeed have a true peak reading watt meter. The 989D's peak and hold detector circuit does require an external power supply or an internal battery. Accuracy? My AL-80B's meter indicates about 5% less than the 989D. Both have the same basic circuit design.
  2. K7JEM

    K7JEM Ham Member QRZ Page

    I think Steve's method of checking the meter is pretty good. The PEP of a CW dit should be the same as the average power of a key down CW test. If it is not, then the meter can't possibly be reading a true PEP in any mode. The peak meters have a diode that charge a capacitor for a short hold at whatever that peak value is, doesn't matter if it is SSB, CW or AM, it ought to detect and hold (at least for a few hundred ms) the peak reading.

    The C model of this tuner has an inferior design in the peak reading circuit. The D model looks to be much better, and is probably far more accurate on peak reading than the C model.

  3. K7JEM

    K7JEM Ham Member QRZ Page

    I am describing the circuits actually used in the wattmeter in question, and the 989D model which uses some active components to improve the peak holding capability. You can look at the circuits yourself, if you are so inclined, or you can continue to insult people in the thread. That choice is yours to make.

  4. G0HZU

    G0HZU QRZ Member

    The old PEP Wattmeter debate again...
    In reality I think there are several mechanisms at work that make it hard to measure PEP using regular test tones or dits or dahs. The radio's amplifier will have an ALC circuit with a specific attack and decay time constant and this can mean the amplifier can produce a different (ALC limited) PEP on a scope depending on the duty cycle of the waveform or its shape. Basically the ALC can be too slow to catch some of the waveform peaks so not all waveforms are ALC limited.

    In my limited experience of tinkering with this subject (limited because I got to the conclusion pretty quickly) I find that the only effective way of measuring true PEP of a Ham or CB radio is to use a calibrated scope via a coupler or attenuator and use your eyes to look for the spiky peak voltage of typical speech.

    This is because there will be parts of the human speech waveform that can give spiky speech peaks that are too fast for the ALC (or a cheap PEP meter) to respond to. So the true PEP will be different to what you would get with a regular two tone test signal or regular pattern cw signals. I'd expect the ALC to cope differently for all these waveforms so the amplifier will deliver different PEP in each case.

    If there are regular peak/hold meters that can claim to cope with human speech then I haven't seen one yet. But that doesn't mean they don't exist. I simply haven't looked hard enough but I'm a believer in simply using a scope anyway :)
  5. KV6O

    KV6O Ham Member QRZ Page

    Peak = PEP? Peak = Peak Envelope Power?

    Peak power is the product of the peak voltage and the peak current. These are instantaneous values. Peak Envelope Power as defined by the ITU and the FCC is the average power supplied to the transmission line by the transmitter during one RF cycle at the crest of the modulation envelope. Find the Peak voltage, compute the average - PEP = (PEV x 0.707)^2/R

    I don't know what the meters on the market are doing, but if it say's Peak, I would think it means Peak, not PEP, although I could be wrong! But that's the question being posed here, and exactly what KQ4X was asking. I use a LP-100, and the instructions don't say mention anything about PEP, only Peak and Average. It would appear that PEP and Peak are being used interchangeably out there in the watt-meter market.

    So, it looks like the question remains - are the meters out there showing us peak power, or peak envelope power - which are two different things.

  6. G0HZU

    G0HZU QRZ Member

    For me power is best defined as the rate of doing work so there has to be a defined measurement time involved.

    For average power:
    You make the time very, very long wrt the period of the lowest modulation frequency and just measure average power (average rate of work) using
    something slow like a thermistor or thermocouple meter for the best result.

    For PEP:
    You try and calculate the average power at the modulation crests inside the defined time frame of 1 RF cycle and this is the PEP. The usual
    approach used by hams for determining PEP is to simply use a peak RF voltage detector and calculate the PEP as below.
    PEP = (peak voltage * peak voltage)/100 for a 50 ohm system.

    In theory at least, you would actually need to measure the true rms voltage within the timeslot of 1RF cycle and this isn't practical for a basic and
    cheap metering system. Besides, as long as you assume the highest frequency waveform is the RF signal itself by a significant margin (i.e.
    harmonics and other higher frequency spurious terms are well suppressed) then the traditional method of detecting peak voltage is going to give
    a good result for the PEP calculation. This is because you can reliably calculate the true rms voltage from the peak voltage if you assume the waveform is a pure sine wave.

    But you have to be able to capture and measure those spiky peaks accurately and most of the cheapo diode based peak/hold meters I have used
    fail to do this by some margin so I prefer the traditional scope method.

    In short, you just want to know the highest 'rate of work' of your 14MHz signal (i.e. at the modulation crests) and so surely this needs to be defined over all
    of the period of one RF cycle to measure/prove the 'rate of doing work' at the required transmit frequency? That will be why the PEP definition
    dictates the measurement time as 1 RF cycle.

    If you make the measurement period even shorter than this and try and compute average power across just a degree or two of the RF waveform
    itself right at its voltage peak (as if you treated the peak voltage as a dc voltage) then I'm not sure where it takes you in terms of finding anything
    meaningful. But maybe a few CB amplifier advertisers would find something LOL!
  7. K7JEM

    K7JEM Ham Member QRZ Page

    The manual says PEP, which would seem to be correct, if the meter can process and hold the reading for a very short duration, such as one "dit" as Steve mentions.

    PEP and average should be the same on these meters, which they apparently are, on continuous (or long) transmissions on CW or FM. The only real question is whether or not the meter will actually respond correctly to the very short peaks you get when using SSB, or a string of morse code characters. If you have a 100 watt output transciever, you should see 100 watts on the meter in the peak position, whether you are sending CW or SSB. In the average position, CW would be somewhere around 50 watts (depending on character string and speed), and SSB might be around 25-40 watts, depending on audio characteristics and compression, etc.

  8. KQ4X

    KQ4X Ham Member QRZ Page

    I'm reading all this guys! Appreciate all the analysis.

    So far I'm of the opinion that the meter on my tuner, which says "PEAK" is just that: Peak of the signal, and NOT a display of the actual Peak Envelope Power (when in the SSB mode). Which is what my opinion has been all along. After all, if these meter makers KNEW their instrument was capable of determining and displaying true PEP then instead of the meter buttons reading "PEAK" they'd all say "PEP" or even spell it out as "Peak Env. Pwr" or such.

    Now, my frustration is with the FCC, not with the power meter makers (like my MFJ tuner). The FCC acts like it does not know how exotic and elusive Peak Envelope Power is! We would be better off for them to go back to designating maximum power INPUT to the final amp stages, as before. Damn, things were so much easier then.

    One other comment: We can't expect manufacturers to install an accurate and equally expensive PEP meter on their products. It would be cost prohibitive, and we wouldn't want to pay that price.

    Enjoying the conversation.
  9. G0HZU

    G0HZU QRZ Member

    Hi KQ4X
    I'm not sure I agree with all of your conclusions here because I think most ham meters will be scaled to try and indicate PEP. Your meter might be
    different but I'd be quite surprised if it was... :)

    For sure most meters will be detecting peak RF voltage but you can still scale the meter in PEP with reasonable accuracy if the transmitter is not
    generating high harmonic levels.

    How they do this is to (try and) detect and hold the peak RF voltage with a peak detector in the way in which we are all familiar.
    This measures the peak voltage of the waveform.

    How they make the leap to PEP is based on the assumption that the RF cycle at the modulation crest is a sinewave with no harmonic distortion etc.
    i.e. if you could zoom in on that single cycle of RF at the SSB modulation crest on a scope you would hope to see a pure sinewave at a frequency
    that is within a few kHz of the suppressed carrier frequency. So you can calculate the true rms voltage of that sinewave by detecting and holding
    the peak voltage and doing a few sums.

    So they can simply scale the meter to show PEP based on the measured peak RF voltage and this all assumes the waveform is a sine wave.
    PEP can then be predicted and displayed using the simplified equation of PEP=(Vpeak*Vpeak)/100 for a 50 ohm system.

    If there are highish harmonics present then the accuracy of the PEP estimation falls away alarmingly because you can't estimate the true rms voltage of a
    distorted sinewave waveform based only on its peak voltage. You can get big errors especially if the phase of the harmonic is unknown.

    Like I said before you really need to measure true rms voltage in that one RF cycle and PEP = (Vrms *Vrms)/50 for a 50 ohm system. But measuring
    true rms in this tiny timeslot is out of the reach of a cheap and basic meter.

    If you look at both equations above they both are a (voltage*voltage)/resistance so the only thing you have to do to get PEP indicated on a meter
    that detects peak RF voltage is change the scaling on the meter face to suit PEP. (assuming the RF waveform is a sinewave as described above)
    So on a healthy radio with low harmonics (eg typically better than -46dBc) the estimation of PEP based on peak voltage readings should be
    accurate enough for most people.

    Bird recommend the radio has harmonics better than -50dBc if you don't want to impact the accuracy of some of their power meters that detect
    peak voltage as a means of determining RF power.
  10. G0HZU

    G0HZU QRZ Member

    I should also point out that most meters will already have the relevant scaling change done because even on the cw setting the meter will need to be scaled to indicate power based on peak RF voltage measured. i.e. the meter scaling needs to have the conversion of Vpeak to Vrms factored into the scaling.

    So even in cw mode the meter will indicate power based on Power = (Vpeak*Vpeak)/100 for a 50 ohm system.

    On the SSB/PEP setting you just need active circuitry with a very fast attack time to detect the rapid and spiky peaks and charge the hold capacitor and then a buffer amp to drive the meter without discharging the hold capacitor too quickly.

    The actual scaling of the meter should be the same as in cw mode. Just the dynamics of the detector are changing, NOT the scaling. That's why I seriously doubt that there will be meters that aren't scaled in PEP when they try and do the peak hold function in SSB mode.
Thread Status:
Not open for further replies.

Share This Page