EP-AB003 linearity

  • Hi all, I finally got the EP-EB003, it would appear that the main regulator is not OK. First the 2 tone drive signal from an IFR2026 signal generator and an F1OPA driver amplifier, 14dBm per tone. Nice and clean. Sorry for the photos, in future I will grab the trace from the spectrum analyzer.

    Now 4W PEP from the EP-EB003, 30dBm (1W) per tone

    -35dBc 3rd order referenced to the tone, 20kHz span, not as good as I would like put probably just about acceptable. But look how the IMD comes back up. That is not right of course. So now look at 100kHz span, very ugly and it looks as though the regulator is oscillating.


    These sidebands are around -40dBc so they would in the noise for most people but I would never radiate such a signal. Reducing the drive level does not improve the situation.

    Does anybody know what the voltage regulator device is in these amplifiers? I would have had a poke around but could not find a suitable Allen key to take the lid off. Of course I have one.... somewhere.

    The power supply used for testing was a big linear supply and as you can see from the first photo the drive was clean. This comes from the PA.

    Has anyone else got an EP-AB003 that they can check? You really should!


    Conrad PA5Y

  • Hello Thomas, each measurement was done for a few seconds only, the problem that I see is not thermal. If I can cure the instability then I will add a heat sink and a blower as it is good engineering.

    As I said these spurs are 40dB down on the main signal and so would most likely be inaudible on the transponder. It could be something simple like a dry joint on a s capacitor on my example. I would very much like it if someone else could check another amplifier on a spectrum analyzer.

    Best Regards

    Conrad PA5Y

  • Does anyone have a datasheet for the YP 242084 amplifier and the YS 1802 regulator used in this PA?

    I know only Innotion YP242034:




  • Thanks Sandro.

    I had a look under a microscope and one device has YP242034 printed on it and the other YP242084, as the ....84 does not exist it seems that this is a printing defect. I am going to connect an external 6V supply next. I checked for dry joints etc and there was nothing. It appears that I have a 15kHz oscillation somewhere.

    There are many of these amplifiers in use so I am surprised that nobody has measured the 2 tone IMD before. I hope that I have an isolated case.


    Conrad PA5Y

  • I found the problem. I looked at the main DC and bias supplies and I noticed that there were square wave pulses at the same frequency as the tone separation.

    1. I disconnected the main supply and fed 6V externally via the 2 BLMs, the problem was still there.
    2. I realized that RX/TX changeover was occurring at the same frequency as the tone separation
    3. As soon as I shorted pins 4 and 5 on the IC TPH2504 the problem went away

    It was interesting that this behavior occurred at all drive levels so I suspect that there was a problem with the input power detector, I did not investigate further as I always intended to run the amplifier in TX mode. In order to keep things thermally stable I will use a FET switch to turn on the 6V supply when it TX. Next I will try with the internal SMPSU again, it is probably fine.

    On my unit the input attenuator is not populated and the gain is 16dB. 1dB compression occurs at 4W PEP. The IMD5 does not always reduce at reduced drive but the general trend is OK. This PA is fine for use on the transponder at 4W PEP, or 1W per tone for 2 tone testing.

    Tone spacing 1.2kHz, 6V DC fed from bench PSU

    Drive per tone (dBm)PEPIMD3 ref toneIMD5 ref toneIMD7 ref tone

  • Hello Thomas, yes of course it is. That does not mean that you cannot use an external 12V supply. I don't like these 12V AC/DC plug in SMPSUs they radiate noise on the lower bands. I operate on 144MHz EME and such 'wall warts' are noisy and I never use them in my shack. I have 12V linear supplies in the shack instead.

    The internal regulator reduces the 12V DC down to 6V for the PA, there is also another regulator to reduce the 6V in order to generate the necessary bias supplies for the YP-242034 PA ICs.

    If you are relying on RF sensing for RX/TX switching then you might want to check that you don't have these sidebands.

    It could also be that my unit was defective, the easiest solution was to short pins 4 and 5 on the IC TPH2504 and hence disable RX/TX switching. Now the performance of this amplifier is quite reasonable, especially considering the price.


    Conrad PA5Y

    • Official Post
  • Hello Thomas.

    Of course I have read that thread. My amplifier has 16dB gain as I have indicated in my posts above. One obvious difference is that the attenuator is not populated on my amplifier. Instead there is a 0 ohm resistor for the series element and the shunt elements are not populated. It was delivered in an EDUP box and it appears to be genuine but who knows.

    Or are you suggesting that I do not know how to measure the gain of an amplifier :)


    Conrad PA5Y

  • Hello Thomas, I go on vacation next week but hope to have full duplex ready upon my return. Currently I only have RX on the dish but I have a POTY ready to go, I am looking forward to our first QSO. I have a 1,1m Triax dish.

    I hope that my experience with this amplifier is useful to others.


    Conrad PA5Y

  • Having thought about this the RF detector was working perfectly well. A 2 tone signal looks like 100% AM so of course the detector switches off at the zero crossings, this would never happen with an SSB/CW or FT8 signal. Only a 2 tone test would cause this behaviour.

    I feel foolish :)


    Conrad PA5Y