Amplifier input impedance fault

[TheWookie]

Hi, I'm new to the forum. There seems to be a lot of people here with experience of RF power amplifiers and I was hoping someone might be able to give me some pointers.

I have an amplifier which is not working properly. With 10W going in it should output 125W (as compared to a working amp) this one isn't, it's only putting out around 20-50W. One thing I noticed is that the input SWR is very high, like 5:1. With 1W going in the SWR is ok, but at 10W input swr is very high. I find it strange that the input swr increases as the input power increases. For reference, input swr on a working amp is 1.5:1 with the same 10W applied.

Both output transistors have been replaced (old ones confirmed dead), switching relay has been replaced. Bias and collector voltages are present. I've pulled out and tested every single other semiconductor in the circuit. With reference to the circuit diagram (attached) I removed C1 and D10 which isolated everything except the amplifier part. Still get the same problem.

I'm at a loss now as how to even fault find it. it looks like a problem somewhere with the matching circuit perhaps ?

 

[Crusher]

Did you put the same brand and type transistors in?

 

[Shockwave]

You might want to check on an antenna to see if your output VSWR is skyrocketing at the same time? If it is, chances are the amp is going into an oscillation at the higher drive power. Since the output transistors have already failed, that condition may have overloaded the DC chokes and bypass caps feeding the output transistors.

These are chokes L1 through L4 and caps C30 through C36. In particular, look for any discoloration of the chokes. If they shows signs of overheating, they have lost some of their ability to decouple the RF from the DC and could be a source of RF feedback at higher power. One common path for this to occur is when enough RF backs up into the DC line and finds it way into the bias circuit. VHF amps are more sensitive to these issues.

 

[nomadradio]

A ten-to-one power gain is more than I'm used to seeing from just one push-pull pair of transistors on 2 meters. Found a review on eham.net that said more like six-to-one for this model. Much closer to what I'm used to seeing from bipolar transistors.

Detaching C1 should serve only to disable the keying circuit. Leaving it connected should boost the input-side SWR only a very little. It is meant only to steal a tiny bit of the RF-drive power to activate the relay, not enough to increase the mismatch.

Just as an educated guess it sounds like a failed capacitor. Not easy to check in the circuit. I would start by comparing the temperature of the two RF power transistors. If one of them runs hotter than the other, look for a bad cap feeding into (or out of) the transistor that's running cooler.

This kind of amplifier circuit is not well isolated between input and output. Any mismatch trouble in the output will disrupt the input impedance.

73

 

[Shockwave]

A ten-to-one power gain is more than I'm used to seeing from just one push-pull pair of transistors on 2 meters. Found a review on eham.net that said more like six-to-one for this model. Much closer to what I'm used to seeing from bipolar transistors.

Excellent point. The data sheet confirms the Toshiba 2SC2782 operates with a typical gain of 6.8db and requires 18 watts of drive to produce 90 watts output, per device.

 

[TheWookie]

Thanks for the replies.

I made sure to replace the failed transistors with the same make and model (Toshiba 2SC2782) and I made sure to buy a matched pair too.

Only reason I detatched C1 was just to isolate that part of the circuit and eliminate it from being the cause of the problem. Hopefully it narrowed down the problem area a bit.

I have been driving this into a high power dummy load, I did measure the output VSWR, can't remember if it was skyhigh or not on the output. Nothing looks to be overheated or burnt out.


So it's looking like there is a failed cap somewhere. I didn't think to check the temp of the transistors, I will do that and see if one is getting hotter than the other. I didn't want to keep running it in fault condition in case it damages the new replacement output transistors. This amp is designed to take up to 50W input so I guess testing at 10W is ok ?

Regards the gain, I tested a working unit on the bench with a signal generator and low pass filter into a 50ohm dummy load. I put an insertion type power sensor on the input and output. With 10W in it put 125W out. At 50W in it made 195W out. From what I have read these amps seem a bit variable in quality, some people get 200W out with only 30W input.

 

[TheWookie]

I do have a bag of replacement capacitors, hopefully I've bought the right specification ones. Looks like I'm going to have to try them one at a time and see which one (if any has failed)

I don't have any replacement chokes to try, not sure what the correct values are as no values are given in the circuit information. I will try and put a link to some pictures.

 

[Shockwave]

Regards the gain, I tested a working unit on the bench with a signal generator and low pass filter into a 50ohm dummy load. I put an insertion type power sensor on the input and output. With 10W in it put 125W out. At 50W in it made 195W out. From what I have read these amps seem a bit variable in quality, some people get 200W out with only 30W input.

The reason for testing on a resonant antenna is that the dummy load will perfectly hide the exact problem we are looking for. That is any signal content that is not on the fundamental frequency and may be outside of the antennas bandwidth.

 

[TheWookie]

Hi, Just posting a follow up to this. Looks like I have got this amp up and running again now thanks to the posts on here.

I replaced the caps C30 through C36 and that seems to have worked. I get 1.5:1 on the input now (which is consistent with the known good unit). And 105W on the output, which is a little lower than the known good unit, but probably just needs tweaking a bit now to make it happy.

There is a pot on the input to the amplifier stage which affects the input impedance. And also a pot on the output of the amplifier stage. I'm guessing it's this one I need to tweak a little for best power output.

 

[TheWookie]

Ok best I can get it to is

Input = 8.7W @ 1.4:1swr
Output = 93W @ 1.06:1swr

It does peak at 101W initially but then drops to 93W

The new one I tested was about 125W output using the same rig to drive it.