I used the 817 with the amp without any protection without a problem so that diode is doing the job! After doing a bit more reading I added a diode when using it with the KX2 as there seems to be some stories of the lack of protection against back emf causing an issue with an Elecraft rig, but as the reason for buying a KX2 was to reduce the weight of the pack I now tend to just use the KX2 barefoot with an internal battery which is a lot lighter than carrying the amp and battery.
Finally when driving the amp with 2.5w from the 817 I have had no reports of poor audio with an additional (sotabeams) speech processor - harsher audio yes so the processor only gets turned on when it is difficult - and by using a Web SDR when trying it at home I’m reasonably confident it is a clean signal. Paul
This device is rf sensitive. In some (antenna) situations to much rf is coupled into this compressor (via the cables). In some situations I had audio distortions with only 5 W out. When changing the antenna and hence reduce the radiation into the station it was fine.
73, Ludwig
… makes sense I was using it with a linked dipole so not a lot of RF near the processor. I’ve not tried it with an end fed but that could well be a problem!..
Mine doesn’t like 80m from the 80m dipole or the trapped endfed I have.
I had one, seemed to work fine with my FT-818, I did however purchase a PTT cable from a reputable UK cable manufacturer. Using it with the 818 and sotabeams Band Hopper IV working French Guyana whilst doing an F+F activation portable. In fact plenty of enough power to easily work both South and North America from the UK. I managed to blow mine, make sure the first thing you do is add power poles!
However, this little upset allowed someone to investigate the insides of the power amp and the quality is appalling!
The amount of heatsink compound on the MOSFET…no way near enough!
You may also want to consider the PTT mod to protect your rig
If both the heatsink and the MOSFET plate are nearly exact plane, you need only a tininess of compound. More would be disadvantageous.
73, Ludwig
Hi AJ,
I sometimes use one with my FT-818. I use separate LiFePO4 4200mAh batteries for the radio and amp.
I use the amp on SOTA and POTA activations for SSB and at home on SSB/FT8/FT4 etc. Outside it doesn’t usually get too hot, but in the car/van or at home, constant use means it gets very hot. I’ve modified an old PC case fan to sit on top of the amp for these situations.
Mine was new and came with the cable for the FT-818. I have since purchased a cable from TechnoFix in the UK that allows me to PTT the amp and the radio via a USB cable when playing data modes.
I have a frequent recurring issue with this amp/radio combo though. On 40m SSB, I regularly get reports of distorted audio when running the amp on full power out. At 1w in/10w out, it isn’t so much of a problem, but on 2.5/5w SSB input from the 818 I will get complaints. The dipole is matched fine, I’ve attached all sorts of ferrite rings/cable filters, to no avail. Sometimes moving the cables around/shifting the amp away from the radio improves things, but it’s very ad-hoc.
Always 40m SSB for audio complaints and I have also had one complaint of a wide signal on 17m SSB when running full power output. I don’t have a spectrum analyser to look at the output signal, but I can imagine it’s not as clean as it perhaps could be… 
Quite frustrating really, as otherwise the amp is pretty reliable. Having said that, as I tend to only use it on POTA activations, I’ve decided to get myself an FT-891 and will probably sell the amp once my wife Nic passes her intermediate, as we’ll be able to use 50w then in challenging locations without the need to carry patch leads, the amp, the second battery, various connectors etc etc.
73, Simon.
Thanks everyone for the responses, this has given me a lot of useful info and things to think about. This is a great helpful group!
Only saying what the person who looked at it said, the actual component was difficult to find as it was on the underside of the board
There is also this
Sweet baby Jesus another ham who loves the sound of his own voice. 6m30 before we get to see the issue. I don’t have enough minutes left in my life to waste them on these videos that move slower than a snail on Mogadons.
I have also had complaints about distorted audio even on 2.5W input on 40 m and 15 m band, but then other operators have said ignore that guy your audio is fine!
I am not sure how much credibility I would give that schematic in the video or some of the information presented. I have only had my amp for about a month, and it may have circuitry not present in older amps. I drive mine with an IC-705 with no keying problems. I believe that the schematic is more of a simplified schematic or functional block diagram. He was concerned about the relay current and no back EMF protection. The schematic shows the 2 relays connected to the supply voltage and the keying line. He states they are 700 ohm coils with 2 in parallel resulting in 350 ohms. That should result in about 40 ma of keying current. Mine measures .56 ma total. With the power disconnected and the power switch on you should be able to measure this resistance from the positive supply to the keying line at 350 ohms. On my amp this is high impedance. If you check the keying line to ground, there appears to be a diode junction of the correct polarity between these lines. There appears to be additional keying circuitry driving these relays.
He also states the power switch only disables power to the relays and LEDs and the rest of the circuit is always powered. I find this hard to believe. My unit draws no current when the power switch is off. The 78L06 regulator that biases the transistors draws typically 4 ma plus its output current that is driving the bias. The quiescent current of the 2 output transistors is probably 100 to 200 or 300 ma. With the power on the current draw is about 22 ma. With the input terminated in 50 ohms and the output connected to a dummy load I keyed the amp and the supply current increased to about 330 ma and no measured outputs from the amp. That should be the power to the band select relays, keying relay, the LEDs and the quiescent current thru the transistors. The NXP data sheet shows quiescent current in their test amp at 200 to 800 ma.
This is for my amp, your mileage may vary. There may have been improvements made over time. My amp is only a month or 2 old. A great feature of this amp is the low standby current draw when operating on battery power.
73
Burke
KF7NP
Did you mean 56 mA or 0.56 mA?
The high impedance between keying line and positive supply does suggest there is a keying transistor in there now.
I haven’t done it yet, but if there is no diode to suppress the negative spike, the spike should be visible on a CRO.
73 Andrew VK1DA/VK2DA
Andrew
.56 ma measured on a 20-ma scale. NOT 56 ma.
If the relays were driven directly and both had 700 ohm coils the resistance from the positive supply (13.8 V supply) to the keying line should measure as 350 ohms. It measures as open (greater than several meg ohms). There is also a diode junction with the correct polarity from the keying line to ground. This could be a diode or the base junction of a bipolar transistor.
73 Burke
KF7NP
It’s most likely the circuit has developed over the few years it has been in production.