EPT360015C Connex cx3400hp dual mosfet issue.

[Handy Andy]

Ok, I see you're running into a similar problem I had.

Cold Solder joints from their "Silver soldering" alloys used - but then too you did do a much better job looks like the Driver Bias is the one running high, which may have been the CAUSE the ORIGINAL condition.

Locate R219, I see it on the board by the Driver BIAS trimmer - what is it's value?

Looks like the value may be too low, allows more power to flow into the circuit than it should.

It looks like 10K - it will have to be MUCH higher (larger ohmic) value than that to keep the BIAS at a safe current level.

​

Change R219 to a HIGHER value resistor - like 47K to as much as 100K - to help the trimmer properly adjust the voltage - I'd recommend 100K so you can set Trimmer to Upper 2/3rds of it's "Idle" range - less chance of thermal drift at this drive level.

If it's 47K then you're ok, but it's the 6V and above is what will have to be fixed...

Got a 68K up to a 100K spare?

Do this...because 6 volts is just too high of a "default" without some protection...you're going to add a resistor to make this a divider and take away from of that trigger voltage so we don't recreate this problem for later.

.​

 

[Hawkeye351]

R219 is yellow, violet and orange (47k). Change it out to a 100k?

 

[Handy Andy]

Was updating as you read review the above - you can keep the 47K is it's value (it could have drifted off) - need to check real quick to make sure we've solved the Driver blowing up problem - if the resistor was a value not in 20% of tolerance - too much voltage adjust can occur and when someone doesn't know what they're doing - it will be back into your shop soon enough to be placed on the bench and do this all over again.

 

[Hawkeye351]

Andy, I just ohmed out r219 in circuit and it's only showing 860 ohms. Visually it looks new, but could it still be bad or should I pull it out and ohm it again?

 

[Hawkeye351]

I couldn't locate a 47k or 100k, so I used a 55k instead. Now the lowest the gate voltage on the driver pad will go to is 0.08v, highest is 3.61v.

 

[Handy Andy]

KEEP IT!

highest is 3.61v.

WHEW!

[I posted the above quicky so you'd know what to do - lock in load....]

The rest of this is to help you understand why...

The Gate itself is finicky - too much of a good thing being applied to it - turns it into a switch - not an RF amp. So what that means is it will LATCH on and stay on.

Thank you for checking the 47K resistor - we did not know why the Driver failed and its' shown that when you're gate voltage goes above what EVEN THE ZENER will conduct (crowbar) at - you know now, there was something wrong.

Thank you...

 

[Hawkeye351]

So you're saying stick with the 0.08v lowest driver gate and 3.61v highest driver gate?

I do have a 22k I could try instead of the 55k in case the 3.61v needs to be higher, but the low will climb up to probably 0.50v.

What should I do next Andy.
Thank you again, your help is greatly appreciated.

 

[Handy Andy]

Well, the 55K is about as "high" in upper range of Ohmic as allowed ...

Here's a clue as to why...

Your using IRF520's right?
The real numbers you need is buried in the data - the reason you needed to keep the adjustment values low was to prevent a "latching on" moment...

You want the thing to turn on with RF, nothing more.​

Treat the Bias like you'd do with any radio that uses Bias for AB class operation - mostly SSB - - you want it to turn on with a whisper - but not stay on and crowbarring your power supply to death...

3.61 Volts is about the upper limit - you can change the 55K to 47 K if you don't like the results - but something tells me you might want to use that 55K (odd value but ok did you mean 56K (Green Blue Orange)

 

[Hawkeye351]

Actually it reads 55.6k ohms. So I'm guessing either 55k or 56k.

Yes, same colors as you mentioned.

 

[Hawkeye351]

Think I'll stick with this one in.

So, we got the am power adjustment to stop affecting the gate voltage adjustment.

And

We got the bias down to a safe level.

What we moving on to next?

Again, thank you Andy.

 

[Handy Andy]

Put the thing together?

Once all the basics are completed -

Leave All MOSFETS disconnected until Gate Bias can be Preset...

Preset the Bias to EACH to 3.0V - Driver and Final - keep AM Power to a safe 5.6V - now it's time to think about a Wattmeter also keep a Dummy Load on the Antenna jack.

Resolder Leads of MOSFETs' onto their respective pads.

Keep an AMMETER in the Power supply line and check the AM Regulator.- current surges for heating when preparing to power up.

Remember in the next rough - it - in stage - you're going to be adjusting BIAS to get 25 to 50mA for startup.

​

In review I think you are ready to begin the reassembly - but really work on getting the BIAS set for the Driver to see if it will go to about 50mA - leave the Test Points for Final out for now.

Then if possible - unsolder the Gate lead of the Driver. Reconnect Test Point for Finals

- This step is needed - Why?

• - turn off the signal to the Final from the Driver so you can see the pure Idle current for the individual Finals - but remember the Finals sum together because they tap from the same Test point.
• Connect Test Point for Finals

• Why turn off the Driver Gate? Well, you could remove the Test Point Jumper - but if you do that - the GATE will still receive power from the Bias circuit - as well as any RF from upstream at the Predriver section - might not want that affecting the Final adjustments in this case.

Set these for 55mA MAX and 110~115mA between both TOTAL MAX - once set - then return to the Driver and Resolder it's Gate back on.

Make sure to check and recheck your work - ensure Gate for Driver is set, Finals are reconnected - AM Regulator operates, some warmth but not runaway Chernobyl type action occurring.

You should be able to go for standard tune up from this point on - when Ammeter on the power supply leads is showing DRAW but not to the point of where it would be blowing the typical 4~5 amp fuse

 

[Hawkeye351]

Gonna start on that in the morning.

Just 2 questions though:
1. Is there a certain pin soldering order you recommend?
Example: gate pin first, drain second, source 3rd.

2. How do I check the am power voltage for the 5.6v?

 

[Handy Andy]

Since these are set up as GDS - the SOURCE is "ground" so best to start there - with Drain Next then Gate being the Last one - Gates are the most sensitive to static - and the Substrate the Source To Drain is has it's own built in "arrestor" that functions much like a zener - that is the die design itself - and it's PIV (working) Voltage.

How do I check the am power voltage for the 5.6v?

Sorry, I thought you knew...

The Test Points - they are directly from the AM Regulator - AM power adjusts this - and if you have a Front Panel RF POWER knob, best to set it to mid-point and then set AM Power to observe the "ranges" the RF Power knob turns thru.

5.6V to as low as 3V to as high at 12V (possible) for the AM Power range is considered "ok" - it's supposed to be a little more narrower than that from about 4V to as high as 10V - but most users "retune" to "swing" more. I can't help that - except to make them fixed...as in the variables are removed and fixed values are put in their places - might save you and them from a radio blowing up and would allow the radio to last longer but you don't keep customers that way...

​

You test across from AM Regulator Output line say TP9 (if that number was used for the TP of the AM Regulator line from that side - as the case may be) to Ground for Voltage and set it for 5.6V by adjusting AM Power trimmer.

Upon power up - and depending upon efficiency of the Tune up and even the MOSFET's themselves - the radio may produce a carrier when you TX - this is normal and is why we don't want to push the Gate Voltages past 3.0V on initial tests - for they may latch TOO quickly. So keep remembering to Keep the Test Point removed at the ones your NOT checking - yet - then reconnect in steps - AS STEPS - to prevent a runaway condition.

 

[NZ8N]

So, you are telling him to set the idle current to 55mA per device in AM Mode at 5.6v to the Drain? Not on USB or LSB, mike gain all the way off, no signal through the TX path, Drain at around 12v? Isn't this the way to do Idle Current in Bipolar transistors, SSB mode, no Mike gain, no signal? Doesn't the current increase when you increase the input voltage to the Driver and Final(s)?

I guess I didn't pay attention to the model of radio or the modes available to it, being AM/FM only (and PA).
I'd put it in FM, so flat carrier, desolder one of the legs to the predriver or resistors feeding g it,, to disable any or minimize the RF going through.

 

[freecell]

"By the way, I'm using an MFJ 872 Grandmaster watt meter. Reads AVG/PEAK, not pep."

"Was told that the 31 watts I'm seeing on my meter (avg/peak) is actually around 54 watts pep, is that right?"

31W modulated AVG power = 82W PEP


from the mfj872, 873 & 874 instruction manual:

"Forward RF power readings, switchable to indicate either average power or Peak
Envelope Power (PEP) for SSB and AM transmitters."

https://cdn.shopify.com/s/files/1/0289/7782/3843/files/MFJ-872.pdf?v=1586534082

the letters PEP appear next to the red AVG/PEP selector switch.