Re: QMX - smoke - another C107/Q108 failure
I suspect that, as Chris alluded to, Hans has PD7 configured in push-pull rather than open collector, so it's sinking the (minimal) current to drop LIN_REG_EN down to 3.3V when the processor is driving it high.
I'm still confused why we don't see the full supply voltage on LIN_REG_EN before the processor boots up.
It does appear that LIN_REG_EN, once it is pulled down to 0V after the SMPS initialization, stays there.
Jonathan KN6LFB
|
Re: CAT port returns ` ` ` on QCXmini
Actually I just tried it with FLRIG and it has a "Send" function where you type in the string and hit send.? When I did that the QCX returned a single (`).? Also commands like FA; and FA28074000; also returned the same symbol.? So something else is wrong with it.
-Jerry AC5JM
On Tuesday, August 29, 2023 at 09:43:27 PM CDT, Jerry Moyer via groups.io <ac5jm@...> wrote:
Sorry I wasn't clear.? Actually, I was sending the whole command, for example, typing KA; and then hit the enter key but I was getting that ` reply every time I pushed a button on the keyboard, including when I hit the enter key.? Or perhaps I don't understand.? Are you perhaps saying that typing the command and then enter on Putty just won't work because of the way Putty sending the characters before hitting enter?? ?Maybe there's just a setting somewhere that I need to change??
-Jerry AC5JM
On Tuesday, August 29, 2023 at 03:45:44 PM CDT, KEN G4APB <lfoofui.nbz42@...> wrote:
Jerry, you can’t send characters one at a time to the mini, that why you are seeing the ‘ reply. you have to put the whole command together and send it as a block.
73 Ken g4apb?
|
Re: CAT port returns ` ` ` on QCXmini
Sorry I wasn't clear.? Actually, I was sending the whole command, for example, typing KA; and then hit the enter key but I was getting that ` reply every time I pushed a button on the keyboard, including when I hit the enter key.? Or perhaps I don't understand.? Are you perhaps saying that typing the command and then enter on Putty just won't work because of the way Putty sending the characters before hitting enter?? ?Maybe there's just a setting somewhere that I need to change??
-Jerry AC5JM
On Tuesday, August 29, 2023 at 03:45:44 PM CDT, KEN G4APB <lfoofui.nbz42@...> wrote:
Jerry, you can’t send characters one at a time to the mini, that why you are seeing the ‘ reply. you have to put the whole command together and send it as a block.
73 Ken g4apb?
|
Re: QMX - smoke - another C107/Q108 failure
On Tue, Aug 29, 2023 at 02:24 PM, Jonathan Burchmore wrote:
I did a poor job of stating my question. ?I agree with your assessment of the voltage regulator output. ?My question was about the LIN_REG_EN signal. ?If you look at the trace below, channel 2 is LIN_REG_EN as measured on the 3v3 board connector, i.e. downstream of R101 and Q101. ?It is initially around 7.5V, then drops to 3.3V, then to 0 when the SMPS comes online. ?I'm trying to understand the mechanism that pulls LIN_REG_EN FROM ~7.5V to 3.3V. ?Or for that matter why it is initially at 7.5V instead of the supply voltage of 9V.
Secondly, since I was measuring the voltage right at the connector, it seems that the processor pin would also have been exposed........
Jonathan, What you are seeing is the V_IN level when you first push ON or turn ON the Power Supply. This is goes through R101 to the gate of Q101 and LIN_REG_EN. You definitely exceed the 1.7V nominal Gate Threshold of Q101 and it turns on powering up the 78M33. After about 80ms of 3.3V the processor is initialized and comes alive. The processor now drives LIN_REG_EN .... .also....with it's own 3.3V supply (probably has an internal pull-up to its 3.3V). This 3.3V LIN_REG_EN also exceeds the Gate Threshold of Q101 which keeps it ON and keeps the 78M33 going. When the processor decides it's time to enable the SMPS it drops LIN_REG_EN to 0V (or at least <<0.8V ......the Gate threshold voltage of Q101 is not met and it turns OFF turning OFF the 78M33 and turning ON the SMPS. LIN_REG_EN has to STAY at that 0V level to keep the 78M33 turned OFF.? I don't know where the 7.5V you mentioned came from. I also don't know how happy pin PD7 was about being connected to 12V or 9V and then being initialized. 73 Kees K5BCQ
|
Re: QMX - smoke - another C107/Q108 failure
I do remember reading about at least one other person who
killed their QMX by toggling their power supply between 6 and 12
volts, in order to toggle their transmit power. This is similar
to my own scenario, where I forgot to raise the current limit on
my power supply, which went into current limit when I keyed the
radio into a dummy load, then came out of limit when I released
the key, putting a similar voltage transient into the power
supply input. Should I have been more diligent about making sure
my current limit was set correctly? Yes. Is halving and doubling
your input voltage a relatively extreme condition to subject a
power supply to? Also yes. However, simply switching the thing
on is a power supply transient from 0 to 12 volts, so a
transient between 6 and 12 should be well within what it should
be expected to tolerate. I don't think there is a single
DC-powered device in my house that would be harmed by subjecting
it to voltage transients within its specified input range. So I
would have to disagree that this phenomenon should be
categorized as a "constructor error."
Which was exactly my second point.
I don't think there is a single DC-powered device in my house
that would be harmed by subjecting it to voltage transients
within its specified input range. So I would have to
disagree that this phenomenon should be categorized as a
"constructor error."
That you know of... And I didn't categorize it as construction
error, either! :-) It's a condition *I* wouldn't have thought to
address, honestly.
It could still be a problem for an off the self switching supply.
Probably not a problem for a linear regulator, because there's no
window of vulnerability but nowhere near as efficient with the
power consumption.
I refuse to throw stones at this glass house!
Paul -- AI7JR
On 8/29/23 17:40, Stephan Ahonen KE0WVA
wrote:
toggle quoted message
Show quoted text
Those are all good reasons the LTM8078 won't be the best choice
for this particular application, I really just picked it by
going on the Analog Devices web site and filtering for a
dual-output regulator with integrated inductors. There are tons
of parts out there, some more cost effective and/or practical
than others. My basic point is that switching regulators are a
problem that has been solved and packaged into off-the-shelf
components by companies with more engineering resources than the
entire amateur radio community combined. There's very little
reason to roll your own solution unless you have very unique
needs.
I don't think that "it's a single source part" is a strong
argument in the context of a product that contains an ARM CPU,
which I can personally attest to being frustratingly
unobtainable during the semiconductor shortage. I also see an
ADC from Texas Instruments and a DAC from Cirrus Logic, also not
generic parts. At least the power supply boards are modular. If
Analog Devices' foundry burned to the ground and they stopped
manufacturing switching regulators overnight, it's not a huge
deal to just spin up a different power supply board that
occupies the same area and produces the same voltages, without
needing to touch the design of the main PCB.
As far as BOM cost, if every dollar of BOM adds two dollars to
retail price (this is my formula for things I build and sell),
I'd have paid an extra $20 for a QMX with that extra $10 of BOM
in the power supply... And spent several fewer hours
troubleshooting my dead radio. That time is worth so much more
than $20 to me. I actually ended up deciding it was more
worthwhile to spend $100 on another kit than gamble on the
possibility that the same event that killed my IC403 didn't also
damage other parts in less obvious ways, forcing me to spend
even more time (and shipping costs from the local Digikey
warehouse) troubleshooting. Take my $20 instead, please.
>Of all the QMX I have yet seen, other than the Q103/Q104
Drain short (manufacturing problem) I have yet to see a failure
that is not attributable to shorts, damaged components or other
construction errors.
I do remember reading about at least one other person who
killed their QMX by toggling their power supply between 6 and 12
volts, in order to toggle their transmit power. This is similar
to my own scenario, where I forgot to raise the current limit on
my power supply, which went into current limit when I keyed the
radio into a dummy load, then came out of limit when I released
the key, putting a similar voltage transient into the power
supply input. Should I have been more diligent about making sure
my current limit was set correctly? Yes. Is halving and doubling
your input voltage a relatively extreme condition to subject a
power supply to? Also yes. However, simply switching the thing
on is a power supply transient from 0 to 12 volts, so a
transient between 6 and 12 should be well within what it should
be expected to tolerate. I don't think there is a single
DC-powered device in my house that would be harmed by subjecting
it to voltage transients within its specified input range. So I
would have to disagree that this phenomenon should be
categorized as a "constructor error."
|
Re: QMX - smoke - another C107/Q108 failure
The N-Channel FET Q101 will conduct ON if the Gate Threshold voltage is nominally 1.7V. That is with +12V, +9V, or +7.5V feeding the 78M33 regulator. This in turn causes Q102 to conduct and the 78M33 comes alive. This will also?happen if LIN_REG_EN is at the Q101 Gate Threshold voltage? .....a voltage that must be supplied by pin PD7 of the processor after it's been powered up with 3.3V and initialized. I don't know how happy PD7 is with a pull-up to 12V but at least the current is limited by R101. By definition this happens later and it looks like about t=80ms after pushing the ON button. So it looks like LIN_REG_EN from PD7 of the processor just keeps the 78M33 regulator ON until t=250ms causes the switch over to the SMPS. Since PD7 can now control LIN_REG-EN the processor controls the switch timing to the SMPS. I assume PD7 has it's own internal pullup to 3.3V ? ....or is Open Collector
One thing that you want to make sure of is that LIN_REG_EN actually goes below the Gate Threshold of Q101 or <<0.8V or it won't turn OFF.?No "floating" levels allowed, they drift. You won't see the result because the STMP 3.3V is slightly higher than the 78M33 3.3V and D103 is back biased.....but it can still supply current through the "dot OR".?
73 Kees K5BCQ
|
I used three different colors of magnet wire so it was easy to figure out what wire went where.?
--
73, Dan? NM3A
|
Lately I am experimenting with FDT86256 mosfet in my design Class E PA rig and am quite happy with it over BS170.?
Has a higher voltage tolerance and with one mosfet I can get 5 watts.
73
Barb WB2CBA
|
On Tue, Aug 29, 2023 at 11:31 AM, Tony Scaminaci wrote:
Note that the source and drain pins are reversed from the BS170 so you’d flip the TN0606 upside down (flat side against the washer). Others have made this substitution with excellent results. Again, it’s a good idea to place a very thin layer of thermal grease on both sides of the transistors to maximize heat transfer to the board.
For best thermal contact between the round side of the transistor and the board, I recommend , which will better conform to the shape and provide more thermal contact. In this case I'd also probably swap out the steel washer for something more thermally conductive, such as aluminum or copper, since that's what's going to be making the best thermal contact with the transistor. The premium thermal solution would be an aluminum bracket that makes thermal contact with the enclosure, turning the entire thing into a heat sink.
|
Re: QMX - smoke - another C107/Q108 failure
Those are all good reasons the LTM8078 won't be the best choice for this particular application, I really just picked it by going on the Analog Devices web site and filtering for a dual-output regulator with integrated inductors. There are tons of parts out there, some more cost effective and/or practical than others. My basic point is that switching regulators are a problem that has been solved and packaged into off-the-shelf components by companies with more engineering resources than the entire amateur radio community combined. There's very little reason to roll your own solution unless you have very unique needs.
I don't think that "it's a single source part" is a strong argument in the context of a product that contains an ARM CPU, which I can personally attest to being frustratingly unobtainable during the semiconductor shortage. I also see an ADC from Texas Instruments and a DAC from Cirrus Logic, also not generic parts. At least the power supply boards are modular. If Analog Devices' foundry burned to the ground and they stopped manufacturing switching regulators overnight, it's not a huge deal to just spin up a different power supply board that occupies the same area and produces the same voltages, without needing to touch the design of the main PCB.
As far as BOM cost, if every dollar of BOM adds two dollars to retail price (this is my formula for things I build and sell), I'd have paid an extra $20 for a QMX with that extra $10 of BOM in the power supply... And spent several fewer hours troubleshooting my dead radio. That time is worth so much more than $20 to me. I actually ended up deciding it was more worthwhile to spend $100 on another kit than gamble on the possibility that the same event that killed my IC403 didn't also damage other parts in less obvious ways, forcing me to spend even more time (and shipping costs from the local Digikey warehouse) troubleshooting. Take my $20 instead, please.
>Of all the QMX I have yet seen, other than the Q103/Q104 Drain short (manufacturing problem) I have yet to see a failure that is not attributable to shorts, damaged components or other construction errors.
I do remember reading about at least one other person who killed their QMX by toggling their power supply between 6 and 12 volts, in order to toggle their transmit power. This is similar to my own scenario, where I forgot to raise the current limit on my power supply, which went into current limit when I keyed the radio into a dummy load, then came out of limit when I released the key, putting a similar voltage transient into the power supply input. Should I have been more diligent about making sure my current limit was set correctly? Yes. Is halving and doubling your input voltage a relatively extreme condition to subject a power supply to? Also yes. However, simply switching the thing on is a power supply transient from 0 to 12 volts, so a transient between 6 and 12 should be well within what it should be expected to tolerate. I don't think there is a single DC-powered device in my house that would be harmed by subjecting it to voltage transients within its specified input range. So I would have to disagree that this phenomenon should be categorized as a "constructor error."
|
Wayne,
As Tony AD0VC pointed out, the commutating diode across L14 is definitely a good idea regardless ?of which FET you use. The TN0106 produces a much higher spike voltage than the BS170 because it’s turned on much harder. For some reason, the BS170 isn’t very tolerant of voltages exceeding the 60V breakdown voltage while the TN/VN devices seem to be able to tolerate a lot more. Still, the data sheet specs should be strictly honored regardless of what others are reporting.
My suggestion of alternative FETs is to improve reliability. The additional power output is a pleasant side effect.
Wish you all the best with Idalia - stay safe.
Tony - AC9QY
|
This is interesting, Tony. I'm gonna have to study the schematic to determine the direction of the diode. I think I read of this solution in another post.?
For now...I live in the far SE corner of GA, and we are preparing for the arrival of Idalia. Ooph! I can't wait for this to get over with...
|
My apologies to the group. I called out the wrong transistor - I meant to spec the TN0106 as a good replacement for the BS170. Thanks Tony for refreshing my memory, it’s been months since John Z and I investigated this and I got confused by references to the VN0606 others have been trying. And thanks to Bruce for flagging the TN0606 as having a much higher input capacitance which led me back to comparing data sheets.
The TN0106 device has a significantly lower turn-on threshold than the BS170 which results in the PA FETs operating closer to saturation and therefore dissipating less power. As a result, more power is transferred to the LPF instead of being wasted as heat. The TN0106 is more compatible with the 5V gate driver than the BS170 which really needs about 6-7V of gate drive to achieve the same level of efficiency.
Tony - AC9QY
On Tue, Aug 29, 2023 at 6:31 PM mux_folder2001 < canthony15@...> wrote:
I ran a TN0106 in my high band QDX for several months and had no issues. I had better output on 10 meters than with the BS170 but nothing dramatically better. I also ran the VN0606 in my low band unit at the same time and also had no issues and performance
was good. Those tests were with otherwise unmodified QDXs. Now I am running with BS170s again but with a 1N4148 commutating diode across L14. On my scope I could see that the spike at end-of-tx was eliminated with this diode.
My original problem was that I fried three sets of BS170 in a few weeks. Every thing seems OK with the diode or when using the TN0106 or VN0606.
What I could see on my scope with the Bs170 as originally designed was a 70 volt spike, presumably discharging L14 at EOT, that was chopped off at 70 volts apparently due to avalanche. It had a wide flat top. With the TN0106 and VN0606 the spike was up to 120
volts but with a narrower flat top. Also avalanche but at a higher voltage level.?
It appears that the TN0106 and VN0606 both avalanche at a level more like a 100 volt transistor and I suspect that the manufacturer (same for both devices) is selecting the 60 volt devices from runs of the 100 volt devices. Not so for the BS170 which avalanche
at a level consistent with a 60 volt device.
Anyway, I think that the diode across? L14 is a better solution than using the TN0106 or VN0606. Yes, both methods work, but why stress the transistors at all? It seems to me that you have more margin with the diode. Likewise for using Zeners, again it probably
works but you are adding two parts instead of one and adding parts on the drains of your PA devices. Also you are discharging L14 through the output transformer winding which seems sub-optimal to me.?
I have an issue in my antenna system where the SWR goes up when it rains. It has been dry recently but we had rain again a couple of nights ago and I made it a point to get on 20 meters where the SWR was up to 2.5 to 1 (antenna impedance here goes low when
it rains btw) and I ran for an hour with no ill effects so my confidence in the diode approach is building.
Just my perspective on this issue. Based on my experiences thus far with QDX and my 48 years as an EE, I would add the diode. It is cheap protection and is situated where it can't really hurt anything. During transmission, the inductor current is basically
DC and so there is no risk of creating spurs. It is only at the end where the current stops that the diode does anything.
73,
Tony
AD0VC
The TN0606 seems to have a much higher input capacitance than the BS170 and this may translate into slower switching and increased dissipation especially on higher frequencies . ?
Hard to say if they will prove more reliable in that case although the failure modes may be different?
|
QMX -- first try, first problems (bad sweeps, no output power)
Hi!
I received my QMX today, assembled it, and flashed latest firmware
(1_00_009). I didn't test it on air yet (only with a dummy load).
I have several issues with it. Two are serious, other are minor.
First serious issue. Audio filter sweeps look bad (done with a dummy
load). Yellow trace (USB) is at:
20m: -32 dB
30m: -28 dB
40m: -39 dB
60m: -43 dB
80m: -26 dB
Here are the sweeps:
RF filter sweeps peaks aren't great either, but I guess it's related (?)
What also concerns me is that the filter is set too high for 60m and way
too high on 80m. Can it be fixed easily?
For completion, here are my image sweeps. I'm not sure how to interpret
them (read: I have no idea how to interpret them), but still, the upper
(max) level seems way too low.
And here is my band configuration (default):
The second issue is the output power. I don't have an RF meter (at least
not the one I would trust), but I connected an oscilloscope to the dummy
load and measured the peak-to-peak voltage on it. I didn't calculate the
power, because at these values it would be milliwatts anyway...
Dummy load was completely cold after these tries, so I think the readings
are correct.
I also measured current consumption (at 12 V) when transmitting.
20m: Vpp = 268 mV, I = 0.90 A
20m: Vpp = 110 mV, I = 0.48 A
40m: Vpp = 114 mV, I = 1.27 A
60m: Vpp = 80 mV, I = 0.54 A
80m: Vpp = 80 mV, I = 1.17 A
These are two major issues I have and any help or hint where to look would
be appreciated...
Apart from them, I noticed a couple of other things (probably firmware
bugs):
- when using a straight key, tip and ring in configuration is swapped. My
key is connected to the tip, but it does work only when I select "Ring".
Diagnostics shows that it activates "dit" channel
- sidetone volume setting doesn't work
- sidetone frequency setting doesn't work
- battery icon behaves weirdly (it shows full battery even when I went
from 12 V down to 7 V)
Thanks!
|
I ran a TN0106 in my high band QDX for several months and had no issues. I had better output on 10 meters than with the BS170 but nothing dramatically better. I also ran the VN0606 in my low band unit at the same time and also had no issues and performance
was good. Those tests were with otherwise unmodified QDXs. Now I am running with BS170s again but with a 1N4148 commutating diode across L14. On my scope I could see that the spike at end-of-tx was eliminated with this diode.
My original problem was that I fried three sets of BS170 in a few weeks. Every thing seems OK with the diode or when using the TN0106 or VN0606.
What I could see on my scope with the Bs170 as originally designed was a 70 volt spike, presumably discharging L14 at EOT, that was chopped off at 70 volts apparently due to avalanche. It had a wide flat top. With the TN0106 and VN0606 the spike was up to 120
volts but with a narrower flat top. Also avalanche but at a higher voltage level.?
It appears that the TN0106 and VN0606 both avalanche at a level more like a 100 volt transistor and I suspect that the manufacturer (same for both devices) is selecting the 60 volt devices from runs of the 100 volt devices. Not so for the BS170 which avalanche
at a level consistent with a 60 volt device.
Anyway, I think that the diode across? L14 is a better solution than using the TN0106 or VN0606. Yes, both methods work, but why stress the transistors at all? It seems to me that you have more margin with the diode. Likewise for using Zeners, again it probably
works but you are adding two parts instead of one and adding parts on the drains of your PA devices. Also you are discharging L14 through the output transformer winding which seems sub-optimal to me.?
I have an issue in my antenna system where the SWR goes up when it rains. It has been dry recently but we had rain again a couple of nights ago and I made it a point to get on 20 meters where the SWR was up to 2.5 to 1 (antenna impedance here goes low when
it rains btw) and I ran for an hour with no ill effects so my confidence in the diode approach is building.
Just my perspective on this issue. Based on my experiences thus far with QDX and my 48 years as an EE, I would add the diode. It is cheap protection and is situated where it can't really hurt anything. During transmission, the inductor current is basically
DC and so there is no risk of creating spurs. It is only at the end where the current stops that the diode does anything.
73,
Tony
AD0VC
toggle quoted message
Show quoted text
The TN0606 seems to have a much higher input capacitance than the BS170 and this may translate into slower switching and increased dissipation especially on higher frequencies . ?
Hard to say if they will prove more reliable in that case although the failure modes may be different?
|
Lead shape and/or packaging. ? Mike WM4B ?
toggle quoted message
Show quoted text
From: [email protected] < [email protected]> On Behalf Of Wayne Greene Sent: Tuesday, August 29, 2023 7:04 PM To: [email protected]Subject: Re: [QRPLabs] BS170 Fail ? Thank you. Does anyone know if there is a real difference between the variants of the BS170? I just ordered BS170P mosfets. There were other variants, but I went with this one since it was actually available.?
Wayne
|
Thank you. Does anyone know if there is a real difference between the variants of the BS170? I just ordered BS170P mosfets. There were other variants, but I went with this one since it was actually available.?
Wayne
|
The TN0606 seems to have a much higher input capacitance than the BS170 and this may translate into slower switching and increased dissipation especially on higher frequencies . ?
Hard to say if they will prove more reliable in that case although the failure modes may be different?
|
I have a QRPometer from NM0S. (It was originally a 4SQRP kit, and is now offered by its designer.) It is both a power meter and an SWR meter, and when used as an SWR meter has an absorptive bridge
Link:?
On Tue, Aug 29, 2023 at 2:53?PM mike/w1mt < w1mt.qrp@...> wrote: The QRPGuys SWR indicator uses an absorptive bridge in its design. I use it with my QMX. My problem is that I tend to band hop a lot and I often forget to switch it in for tuning. No problems yet, but one day...
-mike/w1mt
|
Re: QMX: Kit Build complete - My observations
Hi,
my QMX is running well at +12V. Had to repair the Q103/Q104 short issue first by moving a bit the bigger one. Fortunately I measured the short before using it.
I went the assembly my way:
- assembling the LCD and command PCB first,
- placing both SMPSs in,
- soldering the DC connector,
- using a +7V PSU current limited at 250mA for the first switch on,
- loading the FW,
Everything went well. Then I assembled the rest. Nice beast.
And yes Hans, if you don't know already, there is really really no place left on that board. What were you thinking? Haha.
73 Bojan S53DZ
|
Paul - AI7JR
Barb WB2CBA
Mike Besemer - WM4B
Bruce Akhurst
Shirley Dulcey KE1L