Topics

V4 TX problems and new finals #ubitx-help

Don - KM4UDX
 

Hello all!  I failed to pay attention and left my beloved V4 uBIX on TX into a high SWR situation with too many watts for too long while and things got sort of too hot. Like Africa hot.

After I got Chernobyl cooled off, every time I hit TX the rig went to ground, so I figured the 510s had gone to heaven. 

And K1 had been sticking randomly for a while. So removed the old K1 relay and put in a new guy. I don't know how to test the new K1 beyond hitting TX and seeing what happens?

I then removed the old 510s, and watched a bunch of uTuber on MOSFET DVM testing.  After testing, one of the two olds were clearly " totally dead". 

So I replaced both - a first for me!  Then adjusted bias to get to 100mA delta on each. So all good.  When I now hit TX with the CW key (I think this is full audio drive into the TX section), the amp draw increase by about 1 amp -- which I think (?) is fine.

Only problem is I get no RF out. When I use the mic input I get the same behavior. No RF. Ditto when I use WSJT-X tune function which triggers the TX and send a big audio tone into the rig.  No RF.

RX is perfect.

What can I check, or what is the likely issue?  I could have messed up the PA repair, or the K1 repair, or something else got whacked in the Chernobyl event that I haven't figured out?
Or am I doing something moronically stupid (which is an actual possibility..just saying...).

And thanks for all for all the help.



pat griffin
 

Check the connection between the traces and the 510s. Check continuity between the 510 pins and the trace not just at the hole. 
73
Aa4pg


On Dec 12, 2019, at 11:47 AM, Don - KM4UDX <dontAy155@...> wrote:

Hello all!  I failed to pay attention and left my beloved V4 uBIX on TX into a high SWR situation with too many watts for too long while and things got sort of too hot. Like Africa hot.

After I got Chernobyl cooled off, every time I hit TX the rig went to ground, so I figured the 510s had gone to heaven. 

And K1 had been sticking randomly for a while. So removed the old K1 relay and put in a new guy. I don't know how to test the new K1 beyond hitting TX and seeing what happens?

I then removed the old 510s, and watched a bunch of uTuber on MOSFET DVM testing.  After testing, one of the two olds were clearly " totally dead". 

So I replaced both - a first for me!  Then adjusted bias to get to 100mA delta on each. So all good.  When I now hit TX with the CW key (I think this is full audio drive into the TX section), the amp draw increase by about 1 amp -- which I think (?) is fine.

Only problem is I get no RF out. When I use the mic input I get the same behavior. No RF. Ditto when I use WSJT-X tune function which triggers the TX and send a big audio tone into the rig.  No RF.

RX is perfect.

What can I check, or what is the likely issue?  I could have messed up the PA repair, or the K1 repair, or something else got whacked in the Chernobyl event that I haven't figured out?
Or am I doing something moronically stupid (which is an actual possibility..just saying...).

And thanks for all for all the help.

<IMG_20191209_215240_01.jpg>

Anthony Gomes
 

Let me try to help you OM with my limited knowledge which I have gained after building my homebrew ubitx based on V5. Disclaimer: I don't own any version of ready made build ubitx :)

My first question will be where you have sourced your replacement IRF510. If it's from reputed source like Digikey or mouser then you can start the troubleshooting or else check if those IRF510 are genuine one.

Since you have replaced both the K1 relay and the MOSFETS at the same time. Check if there is any contact issue with the replaced K1 relay. Since your receiver is working, are you still able to listen to the receiver when to press the CW key ?

As the receiver is working fine I would focus on things located beyond C80.

73
Anthony
VU3JVX

Evan Hand
 

The fact that you could do the bias adjustment implies that K1 is working.  It also implies that both of the IRF510s are OK.  Either could be false, but not likely.

That leaves the drivers from Q90 on. 

To progress from here it would be best to know what test equipment you have on hand.  We will need to trace the signal from the base of Q90 to the gates of the finals.  I find that an oscilloscope works best for me (assuming it can measure the frequencies to be used.

I use the CW mode to test the drivers, as the signal is constant, and easier than doing the single or two tone for SSB. (getting input levels correct can be a hassle) 

I will assume that you have the schematic from HFSignals.

In another thread I posted the waveforms that I measured on the test points up to the finals.  I will look them up if youi have a scope.

73
Evan
AC9TU

Don - KM4UDX
 

I got IFR510 from someplace. I can't track the order history. But using a simple set of DVM-based tests they check out for on/off state. I know that says nothing about the actual performance on the circuit. 

Don - KM4UDX
 

I checked the IRF510 traces per diagram -- red checks indicate continuity with appropriate PA pin.  Looks okay? I now see why a shorted Q94 made the rig short the power supply on TX -- the Source pin goes to ground, and it was shorted internally due to my Chernobyl event. 

Don - KM4UDX
 

Evan and Co.  I opened RV1 all the way and that gave me a small momentary deflection of power meter.  But essentially RV1 setting doesn't appear to be the problem. 

Here is a short vid of RV1 full open and clicking the CW jack.

https://photos.app.goo.gl/X8NDwuquHK65rHoXA

At least nothing really bad happens?

Don - KM4UDX
 

I may get a scope next week. Of course I have  no idea how to use it, but my guru (Dan, W3BU) has kindly agreed to some brief training.

I have a DVM and lots of stick-2-it-tivness.  So they are my tools for now!

Evan Hand
 

Two ways to go forward at this time:

1 - Measure the bias voltages on the transistors
2 - Remove the transistors and do typical diode check on each junction

I would do the bias voltage checks.

You are concerned with Q911,Q912, Q92, Q93, Q96 and Q97.

When I measured the voltages relative to ground I got aproximatly 12.3 volts on Q911 and Q912 collectors, 1.13 on the base and .46 on the emitter.

For Q92, Q93, Q96 and Q97 I measured 11.4 volts on the collectors, 1.12 volts on the bases and .46 on the emitters.

See if any of these transistors have the wrong bias voltages.  That would indicate which ones to pull to test.

Note, the values are approximate.  The important value is the difference between the base and the emitter.  Should be around .7 volts.  Indicates it is conducting properly and not shorted or open.

Only real test that I know is to measure the RF voltages, That is why I use a 100 MHz oscilloscope. 

Try the above and let us know.

73
Evan
AC9TU

Don - KM4UDX
 

Evan - for testing the transistors, do I have the board in normal RX mode?

Jerry Gaffke
 

Don,

In RX mode, relay K1 is not providing 12v to the TX power rail,
so transistors  Q911,Q912, Q92, Q93, Q96 and Q97 are not powered up.

I suggest you disconnect PA-PWR from your power supply so the IRF510 drains
are not getting power.  Then press PTT to put the rig in transmit mode
and measure those voltages for Evan.  That way the IRF510's won't be getting hot, 
and there is no need to put a dummy load on the antenna port.

Some parts of the driver amps might get warm if you hold down PTT too long,
I really don't know.
I'd let up on the PTT switch when not actually looking at the voltmeter.

Jerry


On Thu, Dec 12, 2019 at 07:20 PM, Don - KM4UDX wrote:
Evan - for testing the transistors, do I have the board in normal RX mode?

Evan Hand
 

No, you will need to go into transmit mode.  Unless you have a dummy load, I would disconnect the power from the finals (the brown wire on the 3 wire connector for power to the board).  That way you can hold down the CW key for long periods without actually transmitting.  If you have a dummy load you can use that, just keep the transmissions a reasonable length to not over heat the finals again.

73
Evan
AC9TU

Don - KM4UDX
 

Ah...understood. I have both a dummy load and separate PA power supply, so I can lift that as well. 

Funny that the water heater started cycling last night. These things always go bad at either Thanksgiving or Christmas. Weird. So I'm fixing that first. Then back to the uBITX.

Thanks so very much!!!

I'll report back the results.

Don

 

At 13/12/2019, you wrote:
Two ways to go forward at this time:

1 - Measure the bias voltages on the transistors
2 - Remove the transistors and do typical diode check on each junction

I would do the bias voltage checks.

You are concerned with Q911,Q912, Q92, Q93, Q96 and Q97.

When I measured the voltages relative to ground I got aproximatly 12.3 volts on Q911 and Q912 collectors, 1.13 on the base and .46 on the emitter.

For Q92, Q93, Q96 and Q97 I measured 11.4 volts on the collectors, 1.12 volts on the bases and .46 on the emitters.

See if any of these transistors have the wrong bias voltages. That would indicate which ones to pull to test.

Note, the values are approximate. The important value is the difference between the base and the emitter. Should be around .7 volts. Indicates it is conducting properly and not shorted or open.

Only real test that I know is to measure the RF voltages, That is why I use a 100 MHz oscilloscope.

Try the above and let us know.

73
Evan
AC9TU

Don - KM4UDX
 

ah...this is great! thank you and I'm on it.

Well, given that I've never actually measured voltage on any transistor I'm not totally sure what I'm on. haha.

Don - KM4UDX
 

Evan, Raj, Jerry, + Co.

I feel like I just finished Lab Class for Electronics 101.  My lab notes are attached for credit. 

All the transistors measured with V+12.01 in TX mode w/no modulation and no PA V+ (lift the brown wire).
Results follow your measurements, except three.
1&2:    Q94/95 Gates = 0 vs your 4.14.  When I apply PA V+ (connect brown to +15v), the Drain = 15, Gate =0, Source=0.
Note: I checked U2, and output toggles 0/5 between RX/TX.
3:        Q90 Emitter is 2.8 vs your 3.4.

Anthony Gomes
 

Don, you did a good job. I won't worry about Q90 as you must have seen the document says not to worry about small variation. But I find it unusual that you don't see any Gate voltage for Q94/95 it should be around 4+ volts. Since you are not getting it for both the MOSFET  can you check if there is any voltage at 5v regulator U2 output pin, it should be 5V.

But if I remember correctly you did set the gate voltage to 100mA after changing the IRF510.

73
Anthony
VU3JVX

Don - KM4UDX
 

Thank you!!!! For a Liberal Arts major, that means a lot!!!!

I tested U2 and it toggles between 0 and +5 on RX to TX. And the bias was set at ~100ma per 510.

I did more tests on the IRFs using +15v for PA+

Resting RX,     TX no mod,     TX mod key down

q:94

D  15/15/7+-

Don - KM4UDX
 

opps (sorry, hit enter too soon)

q94

D  15/15/7+-
G  0/0/-.09
S  0/0/0

q95
D  15/15/7->8+-
G  0/0/-.10
S  0/0/0

With key down, the Gates move around a bit. 

Thanks so much for all help here!!!!!!

Anthony Gomes
 

Did you changed the setting of those "Blue" pots RV2 & RV3 after changing the IRF510. Even if you have no worries you could have reduced the gate voltage (remember turning those pots clockwise reduces the voltage/current  and anti clockwise increases them). since the output of U2 is toggling between 0 & 5V while TX your 5v regulator is still alive.

If you have not changed the setting of those pots RV2 & RV3. Then I would suggest to check the voltage reading at both side of R97 & R98.

I feel you are very close to fix your radio, next task will be to fix the Gate voltage/current for those replaced IRF510. Please note I am referring to Gate "voltage/current" but the accurate way is to set the "Current". I personally feel the voltage reference is easy for quick check because for current check you need to set your multimeter in series with your radio power supply.

73
Anthony
VU3JVX