On 10/19/2019 8:45 PM, nm0s_qrp via Groups.Io wrote:

I did a quick check of the power transistors on the main PCB, since the SMT parts are too small to probe safely.  With 12.6v supplied, the transmitter keyed, and no modulation:

            G    D    S
Q3    7.9  9.8  11.7    FQP7P06
Q11   12   0    12      FQP7P06
Q12   5.0  0    0       IRF510

A large deviation from these voltages would indicate damage to the board, or defective component.

Hi Dave,

Thanks for that data, but after comparing with my non-working Nouveau 75 I wonder if you could confirm one of those voltages. If  the drain of Q12 is truly at 0 volts with the PTT keyed, there would be zero resting carrier output, and I'm sure that's not right (even controlled carrier rigs have some resting carrier). I measure 10.5 volts on Q12's drain with the PTT keyed, which produces a resting carrier of 3.25 watts and that seems to be in the right ballpark.

The problem I have is that Q11 gets very hot very fast when any modulation is applied and inadequate modulation - something isn't right in the class G modulator.    Here's what those voltages measure on my Nouveau 75:

          G    D    S
Q3   9.6  10.5  13    FQP7P06
Q11   7.5  0  12.7    FQP7P06
Q12   3.8  10.5  .2   IRF510

This modulator never has worked right.  After the first Q11 melted down, I went ahead and replaced all the big FETs, but that did not fix what's wrong.  The BJTs in the modulator have junctions at least, I don't know of a good way to test the small FETs but at least they aren't shorted.

It looks like something is messing up the bias of Q11 -  if the above voltages help you to point me toward the right area of the modulator circuit, I can make additional measurements and do SMT repair.    I just haven't yet had time to build a spice model for that bootstrapped class G modulator to help me understand how it works.

Thanks and 73,

Bob W9RAN

Hi Bob, got same issue on mine.  Q11 get real hot real fast in tx.  I'm awaiting some FETs to arrive before additional troubleshooting.  Please keep me in the loop as to what you find.  -73-
Jeff W9GY Calumet, Mi (UP)

On 11/1/2019 5:39 AM, W9GY wrote:

Hi Bob, got same issue on mine.  Q11 get real hot real fast in tx.  I'm awaiting some FETs to arrive before additional troubleshooting.  Please keep me in the loop as to what you find. -

Hi Jeff,

You bet.   The only one I've heard was WA8SAJ's and it sounded good thru a linear.

When I first noticed that Q11 was burning up, I saw that the tab of Q12 might have contacted Q11's heatsink.   But while I was waiting for replacement FETs to come I realized the drain of Q11 and Q12 are connected together in the circuit anyway so it wouldn't have mattered if they were touching.

So we both received defective PC boards it seems.   The more I try to understand what's going on with this modulator the more confused I get!    A complete set of voltage measurements from key nodes - each terminal of all active devices at a minimum - would be a big help.

73, Bob W9RAN

In the receive mode, all three of these MOSFETS should have no drive, i.e. the gate voltages should be close to the source voltage.  Bob, I think you have the voltages measured for Q3 and Q11 swapped, as the drain voltage you report for Q3 is the same as the drain for Q12.
It looks like the gate voltages on Q11 and Q12 are not correct.  I would suspect failure of Q10 or Q13.  These would be the ones I would try replacing.

73 Dave

On 11/1/2019 1:30 PM, nm0s_qrp via Groups.Io wrote:

voltages measured for Q3 and Q11 swapped

Ah, no, but I read "IRF-510" and forgot there were two of them, and I was measuring Q1 instead of Q12.   Thus my argument about 0 volts on the drain and resting carrier, etc.

Let me try again with new measured values:

           G        D       S

Q3    9       10.3      12.6    FQP7P06

Q11  7.5    3.5*     12.4    FQP7P06

Q12  5.6    3.5*     2.4       IRF510 (NOT the PA one!)

*  the voltage at the drains of Q11 and Q12 measures 3.5V right after the PTT is keyed, but start to rise rapidly, which creates the thermal runaway problem.   The voltage is > 8 volts within 15 seconds or so and when I measured the temp of the Q11 heatsink it quickly rose to > 50 deg.

As with the values your provided, all the above are with PTT keyed but no modulation.

Sorry for the mixup on Q numbers.   Does this suggest anything new?

73, Bob W9RAN

Dave,

I've got a band-aid that seems to prevent the thermal runaway from happening but I'm hoping you can provide some insight into "why" and perhaps a better solution.

I found some nasty looking oscillations when scoping the gate of Q11 to see why it was turning on - several volts of unstable crud in the 600-800 kHz range.  A 1nF cap to ground reduced the amplitude and shifted the frequency to the 225 kHz range, so I tried higher values.   As much as 10uF reduces the oscillation without noticeable impact as now for the first time I have positive peaks and 100% modulation (even though I see some spiky things on the carrier which I suspect is the same oscillation).

The problem apparently is an instability in the modulator circuit that I'm just minimizing so the circuit can work but and hopefully this will enable you to get a handle on what's really going on. Meanwhile I think I'll try the radio on the air tomorrow morning!

73, Bob W9RAN

Bob, that seems like a good insight.  I have a suspicion that there may be an underlying cause, possibly a manufacturing issue causing the small SMT transistors to leak.  My operating theory is that there may be some solder paste under the SMT parts that is allowing some current leakage.  This would cause Q12 and Q11 to be turned on simultaneously, resulting in the high dissipation you observed, and may cause the oscillations you are seeing.  I'm going to do some checking to see what I can find.

I bypassed the gate of Q11 with a 1 MFD electrolytic as suggested by Bob, W9RAN.  I now get  modulation envelope above and below the unmodulated carrier, which is what I would expect.  However, the device is still running too hot.  More later...
--
Jeff W9GY Calumet, Mi (UP)

I've made some progress on Q11 overheating, and 'crud' in the modulation on this unit.  Various bypass capacitors on Q11 seemed to help, but I didn't think that the FET was cool enough and the envelope didn't seem as clean as it should be.  So I decided to take a look at why the device was running with so much current.  To make a long story short, I paralleled R43 with a 470 ohm resistor. This lowered the drive to Q12 and thus reduced the excessive current through Q11.  Now the transistor runs cool.  There appears to still be sufficient modulation capability in both directions, and the spurious signals now appear to be gone.
--
Jeff W9GY Calumet, Mi (UP)

I will suggest another mod that might be a little easier to pull off, just remove R47.  The key is to reduce the bias on Q12.

I haven’t checked/noticed any heating with my Nouveau 75.  Would you advise this as a routine mod?