开云体育

ctrl + shift + ? for shortcuts
© 2025 开云体育
  1. QRPLabs
  2. Topics

qmx+ current limit on power up after blowing/replacing finals...

 

hi, i have been using my qmx+ for a month or two with success on ft8 on 20 & 10m. everything seemed fine. i have a battery power supply limited at 12v? & current limited to 1.5A
yesterday i decided to try to monitor & possibly send rtty during this weekend's contest. i uncovered the last transmit impediment in this forum (fldigi/modem/tx/transmit-shaping-OFF) and transmitted a few replies. i got 2 confirmations.? then i noticed that the transmit power had dropped from ~4W to about 1.5W and checked other bands - indeed i had blown the finals. i replaced them. then i ran diagnostics - normally - i verified that the transmit power (into dummy load) was back where it used to be. thinking it was solved i went back to the rtty contest and during my first transmit - the power supply went into current limit & drew down the voltage to ~6v.
i opened up the unit & couldnt see/smell anything,? i lowered the supply voltage to 7v & current limit to 150mA. but when i turn on the QMX+ it immediatly goes into current limit. i was able to get to attempt a factory reset, but it didn't fix anything.
so - i am hoping someone has a clue for me in terms of at least diagnosis.?
any ideas greatly appreciated!
thanks
lars.kn5ute
 

Lars,
?
Thats an interesting one.
What band were you on?
Do you know your antenna system's SWR at the operating frequency?
?
On 160m and 80m, the voltages across the finals exceed the Vds spec for the BS170's of 60v,
even when operating into a 50 ohm dummy load.? It gets worse as the SWR increases.
Operating on 40m is also at risk with even a moderate SWR mismatch.
Having transmit-shaping-OFF may cause spikes, making it even worse.
?
So my guess is that you have simply blown your finals again.
?
Good to hear that you can talk to the processor, this doesn't sound too serious.
My best guess is that a 150ma current limit is not enough on startup when running at 7v,
try increasing that to maybe 500ma.
?
At startup, the processor is running on the linear 78M33 regulator, and is drawing
around 130ma from that rail.? Since it's a linear regulator, that's 130ma straight from
your 7v supply.? The processor then starts up the 3.3v and 5v SMPS supplies.
There may be a significant surge filling up the large capacitors at C106 and C107.
Also, the rig will be drawing around 100ma from the 5v SMPS once it comes up at maybe 85% efficiency,
That's 100ma*5v/0.85 = 0.59 Watts from the 7v supply, which is another 0.59w/7v = 84ma.
The rig will be drawing a total of roughly 130+84 = 214ma from the 7v supply, plus a bit more
to deal with that surge.? Hence my suggestion that you increase the current limit up from 150ma.
?
I think the solution to the high Vds seen at the BS170's is to do away with L502.
This will reduce power output, so rewind the 3:2 turns ratio at T501 with the 1:1 turns ratio
used on the 9v build.? Your post gives me reason to get back to it, and do some hardware tests.
Since the 9v case doesn't have quite such high voltages at the BS170's, probably fine to leave it as is.
Here's my old thread on this, not a lot of interest expressed and I had plenty of other things to do:
?
Jerry, KE7ER
?
?
 

Lars,
?
If spikes were partly to blame, L502 may well be the culprit.
If we suddenly shut down the square waves going into the BS170's (and also DAC into the modulator)
at the end of transmission, we still have current flowing in L502 for a few microseconds
that now has nowhere to go, creating a voltage spike across the BS170's.
Yet another reason to transition to a standard CMCD topology by removing L502.
?
Jerry, KE7ER
?
?
On Sat, Mar 15, 2025 at 04:37 PM, Jerry Gaffke wrote:

Having transmit-shaping-OFF may cause spikes, making it even worse.
?
 

On Sat, Mar 15, 2025 at 05:13 PM, Jerry Gaffke wrote:
If we suddenly shut down the square waves going into the BS170's
Jerry,? you may have misinterpreted Lars statement about disabling the shaping.? He disabled the shaping done by fldigi, which conflicts with the shaping done by QMX firmware, causing transmission dropouts in the rtty generated by fldigi. So it has to be turned off for fldigi rtty to work. But the QMX shaping should still be there, eliminating the spikes.? Unless,? of course he also somehow disabled the QMX shaping.??
 

Stan,
?
I think you're right, thanks for the correction.
Was blinded when I thought I saw an obvious case of cause and effect there.
?
Jerry, KE7ER
?
?
?
On Sat, Mar 15, 2025 at 08:14 PM, Stan Dye wrote:

On Sat, Mar 15, 2025 at 05:13 PM, Jerry Gaffke wrote:
If we suddenly shut down the square waves going into the BS170's
Jerry,? you may have misinterpreted Lars statement about disabling the shaping.? He disabled the shaping done by fldigi, which conflicts with the shaping done by QMX firmware, causing transmission dropouts in the rtty generated by fldigi. So it has to be turned off for fldigi rtty to work. But the QMX shaping should still be there, eliminating the spikes.? Unless,? of course he also somehow disabled the QMX shaping.??
 

On Sat, Mar 15, 2025 at 11:20 PM, lars anderson wrote:
i had blown the finals. i replaced them. then i ran diagnostics - normally - i verified that the transmit power (into dummy load) was back where it used to be. thinking it was solved i went back to the rtty contest and during my first transmit - the power supply went into current limit & drew down the voltage to ~6v.
Lars, a broken PA may cause a demage of Q507 the PA voltage regulator. (I saw this some times when other HAM gave me QMX for repair.) In this case the PA voltage is high all the time. In this situation RX may be ok and after replacing the BSS170 TX is working basically. But you don't have envelope shaping. This is dangerous for the PA like others wrote before. Please use Terminal and go to Diagnostics. Without TXing the PA voltage should be 0.x V and not some volt.
For debugging a possible broken Q507 be aware the PA voltage is switched on also in RX mode except you are in Diagnostics mode (or you switched off this feature in a deep configuration menu.)
?
73 Ludwig
 

Thanks for the replies,
clarifications: yes, i only disabled fldigi transmit shaping, not qmx
?
it is still broken but i did some measurements this morning.
1. power off resistance VCC-GND ~3K (steady state), VDD-GND ~2.5K
2. set power supply (battery) no load: 7v 0A, short circuit: 0v 530mA
3. plug in power supply to QMX+: radio off: 7v 0A, radio on: 5.45v 525mA
4. menu comes up: factory reset works
5. measured VIN 5.4v, VDD 3.07v, VCC 4.78v,?
6. diagnostics menu comes up on PC via USB/Putty - screen shot attached.
thanks again,
lars/kn5ute
i still need to digest the technical advice in previous replies, however at this time i am not inclined to alter the circuitry beyond replacing anything i might have fried!
 

one more tidbit - i was admiring my nice screenshot i just uploaded & noticed that the supply voltage shown was greater than 5.45v that the power supply got sucked down to when i powered on the unit.
Power supply voltage/current: measured at the battery.
Putty Terminal Disconnected: 5.45v 525mA
Diagnostics Menu: Main Menu: 5.45v 525mA
Diagnostics Menu: Main Menu/Hardware Tests Submenu: 5.45v 525mA
Diagnostics Menu: Main Menu/Hardware Tests/Diagnostics Subsubmenu: 6.95v 295mA
when i leave the diagnostics sub sub menu, power/current reverts to 5.45v 525mA
this surprised me but i'm sure it's a design feature. i think though that whatever gets turned OFF by the firmware when the Diagnostics sub-submenu is invoked is what is sucking all the extra current.
-la.
 

Lars,
?
The reduction in current to something more reasonable when in a dIagnostics submenu?
sounds like what Ludwig mentioned previously, power to the final is shut down then..
Try lifting one end of L502, see things start behaving better.
?
With all that extra current, something is probably getting hot.
Feel the board thoroughly, see if you can find it.
?
Jerry, KE7ER
?
?
On Sun, Mar 16, 2025 at 09:49 AM, lars anderson wrote:

Diagnostics Menu: Main Menu/Hardware Tests/Diagnostics Subsubmenu: 6.95v 295mA
when i leave the diagnostics sub sub menu, power/current reverts to 5.45v 525mA
 

Hello Lars,
the values from the Diagnostics screen shot are a bit weird. The 3V3 SMPS is working at the limit (Duty cycle = Duty max) but the voltage seems to be ok. The 5V SMPS is working below the limit but the voltage is to low.
?
But at first we will analyze the data from your last post.
?
High current most of the time but not so high if "SubSubmenu" Diagnostics is actice. Like I wrote before the PA voltage is on all the time except Diagnostics is active. So when Diagnostics is active you see 295 mA without PA voltage on. This value is to high but not so high compared to PA voltage is on (525 mA and current limit is active). I guess the high current when PA voltage is on means there is a broken BS170 and/ or a broken driver IC503. Maybe more than one BS170 is broken or a wrong transistor is mounted or mounted in a wrong way. Maybe only a part of IC503 is bad. IC503 is connected to VCC. A broken IC may cause some trouble for VCC (5V).
Please debug following QMX QMX_plus PA problems, broken BS170. To stop the high current go to QMX - Menu - System config - Advanced config! - Mod. high in RX and DISABLE this feature. (But don't forget to ENABLE again later.) This should switch off the PA voltage in RX mode.
?
73 Ludwig
 

Sorry I was writing the same time like Jerry.
 

Ludwig, Jerry, thank you very much.
I could swear I tested this, but evidently not. i pulled the 'new' BS170s out (they're socketed) and the current limit went away.
while they were out, i scoped the gates while pressing T on the diagnostic menu & all had square waves so the 7408 & XOR are ok i figure.
I installed my last set of BS170s, readjusted the current limit on the power supply up to an amp & fired into my dummy load.
it worked. i don't exactly know how to test if Q507 is fried, but a friend of mine says that's the same control point that's used for the reduced voltage setting used for teh SWR sweep & tuning.?
if that's so, i believe it is ok because i could use the front panel swr tuning into my atu-10 & the power shown is about right.
anyway, i apologize for wasting folks time, if i did. this was a learning experience for me, but i think it's ok.
i did go on ft8 and had a 2.3W 10m QSO from Scotland (i'm in texas). phew!
i was worried last night that i might have to buy & build another QMX+, and that i'd better get it ordered before Hans releases SSB since they'll fly off the shelves then.
now today, i'm kinda disappointed i don't get to build another QMX+!
never satisfied,
thanks again
lars/kn5ute
?
 

On Sun, Mar 16, 2025 at 10:39 PM, lars anderson wrote:
i scoped the gates while pressing T on the diagnostic menu & all had square waves so the 7408 & XOR are ok
Lars, square pulses are not enough. You need also 5 V for high. Last time I had a QMX on my work banch with a partially broken IC503. Both outputs had square pulses but one side 5 V for high and the other 4.5 V for high. When sending the IC drew an excessive current.
?
?i don't exactly know how to test if Q507 is fried, but a friend of mine says that's the same control point that's used for the reduced voltage setting used for teh SWR sweep & tuning.?
Yes. But you have to check if the output is reduced using SWR sweep & tuning. A simple way to check Q507 is to use Terminal diagnostics and observe the PA voltage. Without "T" pressed it should be 0.x V and with "T" depressed a voltage a bit below the supply voltage.
?
But nice to read your QMX+ is back to life.
?
73 Ludwig
 

Ludwig,
?
I don't mean to hijack this thread, but I am having a similar issue that I have described in another message. Here is the part relevant to Q507:
?
"Now I'm looking at Q507. I ?have read Ludwig's postings on this. With all the BS170's removed, the Diagnostics screen shows the transmitter voltage at 10.9 volts. Ludwig says it should be close to 0 volts, but maybe that is with good BS170's. The two SMPS look fine. "
?
Does Q507 need to be tested with BS170's installed ?
?
Thanks and 73,
Steve, N2IC
 

Hello Steve, see your original Thread for my answer.
?
73 Ludwig

1 - 15 of 15