Help reclaiming the magic smoke that came out of my 50W PA #help #magicsmoke #troubleshooting


Charles Johnson
 

Greetings all,

In going to adjust the bias for the 40m 50W PA that I've built, I began with R5 turned all the way down as per the instructions. Putting the QXC+, which was outputting 4.24W into a QRP Labs dummy load (supply voltage 13.84V, measured voltage on the load 20V), into Practice Mode, I connected the PA (supply voltage also 13.84V, from the same PSU) in series between the QXC+ and the dummy load. WIth R5 all the way down, I measured 93mA. Turning it up, once I hit 95mA, I nudged it back and the meter read 93mA. Taking the QXC+ out of Practice mode, I then depressed my straight key and measured 4.16V on the dummy load. Unfortunately, I'm pretty sure that the VOM was set to read AC voltage (it defaults to AC), so I have no idea what the actual output was. Since I thought it was DC voltage, though, I decided to have a second go at calibrating the PA.

For my second attempt, it was identical, save I mistakenly did NOT put the QXC+ in Practice mode. I did, though, have a dummy load connected to the amp. With R5 turned all the way down, the multimeter was reading 990mA on key down. I turned R5 up a bit - no change. I then began to turn R5 up some more...and then whatever series of events happened (me looking at the meter and/or applying too much torque on the screwdriver or...), the result was that R5 was turned up quite a bit and some smoke and a burning smell came out from just forward of R5, between it and IC1 (it looked like it was coming from beneath R5 and IC1 looks just normal). Interestingly enough, the position was right near the setting from the previous calibration. Turning R5 all the way down, when I keyed the QXC+, things appeared to work - the PA was drawing 900mA while idle and when I keyed the QXC+, the current draw went up to ~990mA. Beginning to turn R5 up again, things seemed to work...until I hit the (potentiometer) position roughly where the puff of smoke had come out. At that point, current consumption dived down to ~15mA (effectively 0).

As I'm not sure where to start troubleshooting this - nowhere else on the amp seemed to be affected - I would appreciate any and all help. I have not looked at the underside of the PCB, so I do not know if there was any damage there. I can supply pictures of the PCB, if desired, but nothing looks wrong at least on the top side.

Charles Johnson
KF4AYT

PS As the multimeter that I was using to read the current consumption is new to me, I suspect that I had the range set differently during the first measurement. I know that the numbers written down for the second measurement are accurate (ie if I say that the meter read 990mA, I mean that it was showing 0.990 when set to the 10A setting).


Curt wb8yyy
 

Charles

Let's assume that everything with visible damage needs replacement. Designers sometimes do what you did, so we don't have to, to create a design. You obviously took it beyond its limits. After you replace the smoked parts, don't push it so hard. Yes be careful with test equipment. Note the last few watts of RF matter little, not worth the smoke. 

Curt wb8yyy 


Charles Johnson
 

To all,

After some testing - I measured R5 with the amp powered off as I turned the trim pit and then tried the bias adjustment procedure again (QCX+ Practice Mode on) - I have come to the conclusion that R5 needs to be replaced. First, are there any other parts I should replace just for good measure? Second, what is the make/model for R5? (the Digi-Key/Mouser/etc part # would really be appreciated!).

Also, while Hans lists 85mA in the build instruction guide, on my first bias adjustment attempt, I was measuring 93mA, and in another thread, someone mentions a bias value well in excess of 100mA. Is there a common window in which the bias values usually fall?

Charles Johnson
KF4AYT


R. Tyson
 

On Sat, Apr 16, 2022 at 07:50 PM, Charles Johnson wrote:
Also, while Hans lists 85mA in the build instruction guide, on my first bias adjustment attempt, I was measuring 93mA, and in another thread, someone mentions a bias value well in excess of 100mA. Is there a common window in which the bias values usually fall?
My amps were nowhere near 93mA and as mentioned in the instructions your amp may be different... so there doesn't seem to be an ideal value and I think it will vary according to the IRF510's fitted into your particular amp. All you are looking for is the point where they just turn on and that can vary from batch to batch of IRF 510's.  Somewhere else and with a different amplifier there is a tale of woe. The guys involved blew up several IRF510's by setting the bias to a particular voltage or current (can't remember which they were using) then they had a lucky break and found out that IRF510's needed setting individually not to some standard value printed somewhere else. Once they realised that each transistor was different then they stopped blowing them up.

Reg                       G4NFR