Date   
Re: Fix for Spurs and Harmonics on V3 Board

Curt
 

Dale

Yes remnants of many experiments here on this list. Those axicom relays may get your harmonics compliant or very close. Watch pinouts before you put them on the underside, they should work okay on top side. Some added the shielded inductors on the 45 MHz crystal filters, and obtained lower mixer spurs in conjunction with those axicoms. The v5 LC filter did nothing to improve mixer spurs on my v4.

That trap idea was difficult to tune even with a spectrum analyzer, so hence came the shielded inductors.

Curt

Re: Nearly non-existent output power, no mod

Ted
 

On Thu, Sep 12, 2019 at 07:42 PM, Clark Martin wrote:
Use an RF probe or oscilloscope to trace the signal through the relays / filters.
 
 
OK, here we are: 

On 80m and 60m, only the filter between the T/R relay and KT3 shows strong RF.  Power out on 80m is 2 watts CW, 1/2w SSB. 60m shows 1 watt/1/2 watt.

On 40m and 30m,  only the filter at KT3 shows strong RF. Power out on 40m is about 3/4w CW and 1/2w SSB; 30m shows 1w CW.

On 20m and 17m, only the filter at KT2 shows strong RF with power out the best: 2.5w CW but only 1/4w SSB on 20. 17m shows about 1/3 & 1/4 if that....

15m and 10m show strong RF on the filter near KT1, with output on both bands less than 1/4w either mode (or should I just say, at barely detectable limits).



The meter used is the analog dial on a MFJ 949D as my 5W slug is missing from the Bird wattmeter  :(

The two predrivers and 4 drivers are all steel-can 2N2222's with 10-ohm bias resistors.  All six run fairly hot - with heat sinks added - though the 4 more so than the pair. The finals (even reputable ones) certainly get hot enough, too. if i put my fingers on the predrivers or drivers, the RF power (such as it is) decreases somewhat. Q90 is also a 2N2222 can with I believe a 15-ohm bias chip on it.  None of this changed prior to the output failure. 

Obviously, energy is being produced but to what end?  It certainly isn't going into the dummy load or out the aerial/antenna/wirethingy.

More guidance would sure be fantastic.  Thank you, all, for your input.



Ted
K3RTA

Re: Fix for Spurs and Harmonics on V3 Board

Dale Hardin
 

Curt, the pinout on the Axioms is symmetrical across the pins, so they can go on either side of the board.  I'm putting off ordering the toroids and caps for the mods in the hope that someone on the list who has done the mods will have extras available.  I would like to see a photo of the added V5 LC filter to see what is involved.  I am just about finished with some mods of a TenTec Century 21.  I added an AGC, and additional preamp, a Hi-Per-Mite filter (w/o amp) and am working on a digital dial.  With that effort and a long trip in October, I won't get to the uBitx until November.  I probably should have bought a V5 and called it a day, but I really love to build and modify stuff, so a V3 got me.  Up until about four years ago I had a nice HP spectrum analyzer, but got rid of it because I used it so little and it took up too much room.  I really appreciate your comments on this effort.  Regards, Dale 

On Mon, Sep 16, 2019 at 6:44 PM Curt via Groups.Io <wb8yyy=yahoo.com@groups.io> wrote:
Dale

Yes remnants of many experiments here on this list. Those axicom relays may get your harmonics compliant or very close. Watch pinouts before you put them on the underside, they should work okay on top side. Some added the shielded inductors on the 45 MHz crystal filters, and obtained lower mixer spurs in conjunction with those axicoms. The v5 LC filter did nothing to improve mixer spurs on my v4.

That trap idea was difficult to tune even with a spectrum analyzer, so hence came the shielded inductors.

Curt



--
Dale Hardin
24750 State St. Unit 487
Elberta, AL 36530
251-597-9256

--
Dale Hardin, KS4NS
Elberta, AL

Re: Nearly non-existent output power, no mod

Clark Martin
 

Sorry, but you description below is akin to bad driving directions… (turn right at the green house, etc).  

It would be far easier and more accurate to specify where the signal was measured (and how).  
i.e. 
KT2-14 2.6V
using an RF probe and DC voltmeter

check on both sides of the filter (input and output) and on each of the contact pins of the relay in question.



Note: all the resistors connected to the driver and pre-drivers are “bias” resistors.  The 10 ohm would, presumably, be the emitter bias resistor.

You replaced the final transistors, did you readjust the bias pots (RV2, RV3)?


Clark Martin
KK6ISP

On Sep 16, 2019, at 5:49 PM, Ted via Groups.Io <k3rta@...> wrote:

On Thu, Sep 12, 2019 at 07:42 PM, Clark Martin wrote:
Use an RF probe or oscilloscope to trace the signal through the relays / filters.
 
 
OK, here we are: 

On 80m and 60m, only the filter between the T/R relay and KT3 shows strong RF.  Power out on 80m is 2 watts CW, 1/2w SSB. 60m shows 1 watt/1/2 watt.

On 40m and 30m,  only the filter at KT3 shows strong RF. Power out on 40m is about 3/4w CW and 1/2w SSB; 30m shows 1w CW.

On 20m and 17m, only the filter at KT2 shows strong RF with power out the best: 2.5w CW but only 1/4w SSB on 20. 17m shows about 1/3 & 1/4 if that....

15m and 10m show strong RF on the filter near KT1, with output on both bands less than 1/4w either mode (or should I just say, at barely detectable limits).



The meter used is the analog dial on a MFJ 949D as my 5W slug is missing from the Bird wattmeter  :(

The two predrivers and 4 drivers are all steel-can 2N2222's with 10-ohm bias resistors.  All six run fairly hot - with heat sinks added - though the 4 more so than the pair. The finals (even reputable ones) certainly get hot enough, too. if i put my fingers on the predrivers or drivers, the RF power (such as it is) decreases somewhat. Q90 is also a 2N2222 can with I believe a 15-ohm bias chip on it.  None of this changed prior to the output failure. 

Obviously, energy is being produced but to what end?  It certainly isn't going into the dummy load or out the aerial/antenna/wirethingy.

More guidance would sure be fantastic.  Thank you, all, for your input.


Re: Nearly non-existent output power, no mod

Clark Martin
 

Additionally, for the SSB tests what was the AF signal source?


Clark Martin
KK6ISP

On Sep 16, 2019, at 6:39 PM, Clark Martin <kk6isp@...> wrote:

Sorry, but you description below is akin to bad driving directions… (turn right at the green house, etc).  

It would be far easier and more accurate to specify where the signal was measured (and how).  
i.e. 
KT2-14  2.6V
using an RF probe and DC voltmeter

check on both sides of the filter (input and output) and on each of the contact pins of the relay in question.



Note: all the resistors connected to the driver and pre-drivers are “bias” resistors.  The 10 ohm would, presumably, be the emitter bias resistor.

You replaced the final transistors, did you readjust the bias pots (RV2, RV3)?


Re: Calibration #calibration

d.maccallum@...
 

Hello, I just got my rig running and tried tuning it up but can not find any info for ver 5.1 info! tried the write up at the ubitx site don't know for what ver and one for uBitx v4.3 and it kinda worked but the instructions do not match what I am seeing on my rig. by playing with the freq and bfo It's way off.

Re: SDR and Ubitx freq not the same.

Dexter N Muir
 

Make that pot a "proper copper coffee pot"! ... and say it fast when you order it ... :)  Brownie points for Captain Janeway, too.
73
Dex, ZL2DEX

Re: No answer in Antuino group

Ashhar Farhan
 

Sorry, it took a while to get my membership on the antuino group. SBL should work, probably even better than the ADE-1
- f

On Mon, Sep 16, 2019 at 10:04 PM MVS Sarma <mvssarma@...> wrote:
IMHO, no changes needed. Mixer's performance being same, perhaps insersion loss or gain could be a factor. Calibration should tahe care, in my opinion. 

Regards
Sarma vu3zmv

On Mon, 16 Sep 2019, 9:57 pm Rob Bleumer, <rob.bleumer@...> wrote:
Sep 11   

Hi All,
As there is no answer in the ANTUINO group my question here:

As I used SBL-1 instead of ADE-1 from my junk box I wonder if there are differences to make into the software.

They are not exactly the same on specs and I do not know enough on the software.

Can someone tell me if I have to make changes and where?
Attatched photos of my home build ANTUINO.

Thanks

Rob PA0RBL

Re: Digimode low power output

Adrien F4IJA
 

Hi woody. 

Thank you for the schematics. I'll try first to check time but it seems to be OK. After that I could try something like your circuit. 
--
73's
Adrien F4IJA
https://www.qrz.com/db/F4IJA

Re: SDR and Ubitx freq not the same.

Werner de Vos
 

I think I have to explain a little more. My ubitx is calibrated and works on the right frequency. I can oparate FT8 or WSPR without any problem.

The thing is, when I connect a RTL Dongle (hooked up the CEC way, with cat control via SDR console) all works fine exept the RTL-SDR RX frequency is off. So for example. receiving with SDR 14.076MHz is 14.0775. 

It can possibly be the SDR dongle that is somehow off frequency.

Re: Nextion with KA78T05

Murray Wills (ZL2IQ)
 

Hi I did the same only the KA78T05 I got (bought 10 cause they were cheap) went red hot 🥵 so I shut it all down and went back to the step down to power the Nextion. Luckily it didn’t seem to introduce any noise for me.

Something wrong with the KA78T05’s as they should have been able to handle the current.


From: BITX20@groups.io <BITX20@groups.io> on behalf of MVS Sarma via Groups.Io <mvssarma@...>
Sent: Tuesday, September 17, 2019 4:12:57 AM
To: BITX20@groups.io <BITX20@groups.io>
Subject: Re: [BITX20] Nextion with KA78T05
 
I suffered with 10pcs of  bad 78t05 from ali seller and demanded refund and got it.

On Mon, 16 Sep 2019, 9:37 pm Herman Scheper, <Hsrscheper@...> wrote:
I am using the Nextion 5"display. When building my Ubitx V5 I used a small step down converter to generate the 5V for the Nextion display. It all worked 100%
Now the building is almost finished i want to replace the 5V stepdown converter. ( It is giving a little amountt of noise…)

I ordered ( Aliex..) 5 pcs of KA78T05 TO-220 stabilisors. ( is the 3A version-  )

Then I replaced the 7805 mounted on the Arduine and replaced it for the KA78T05 version…. No working Arduino and no working display.

I find input=12V output=0V.   Checked before and after the pinning  1=input / 2 = earth/ 3 =output   Find a datasheet on Internet and can not find what I am doiing wrong.

After that make a test print only for the KA78T05 and I can not get the 5V output. Output also on a 2nd KA78T05 is still 0Volt.   Frustrating.  ( I used a 7809 on the same test and output was 9.00 Volt!)

What I am doiing wrong...…….?     ( ordered today local few 78S05   ( = 1.5A version  )

So who knows what I am doiing wrong...?                       Tnx.  Herman

Re: Nextion with KA78T05

Razvan (M0HZH)
 

Why would you need another regulator ? The LM7805 is designed for 1A output and the 5" Nextion display + Raduino uses about 500mA (per specs).

You only need a heatsink for the LM7805.

This will be true for any other linear replacement you use (KA78T05 / 78S05 etc) as the dissipated power is the same.

Re: Nearly non-existent output power, no mod

Ted
 


Clark,

Please forgive my lack of clarity.  I'll try to fashion up something better and make a go of individual relay measurements within a week or so; I'm about to have vertebrae C3-C7 fused and bracketed together, so, any messages after 18 Sept may be Oxycontin-laced  :O   At least I'll have plenty of time.

My chief focus was to answer the concern of whether the relays were switching, i.e. if RF was being sent through the wrong LPF for all but one band, or through more than one filter at once, etc.  I get the impression that the filters are switching in when they 'sposed to, in country parlance, though I'll grant that this fails to say whether good connections are being made in the process.  Again, I'll rig up a better test bench and do a more precise analysis.

I made the measurements using a simple oscilloscope with a cap on the pointy end of the probe.  It picked up some stray RF on adjacent points but for the nature of the test, I let that pass. 

RE: "Bias" resistors, I'd assumed that people who've been around for a while here know that when switching from 3904's to 2222's in the driver stage, it's the 20-ohm [emitter] bias resistors that get doubled up for a net 10-ohms.  My bad.  On the matter of the 2N2222's, I found that transistor mod very worthwhile. 

Oh, the audio source?   The microphone and a whistle, which used to yield upwards of 10-12 watts on an analog wattmeter dial on 80/40m and round 6-8 watts on 20m. 

I'll report back....


Tnx,

Ted
K3RTA

Re: Nextion with KA78T05

Jerry Gaffke
 

Could be that you got bad parts.

Or could be that the part is getting too hot, and going into thermal shutdown.
So you need a better heatsink.


On page 2 of this Fairchild datasheet for the LM78T05, it says that
the "thermal resistance junction to air" is 65 degrees C per Watt.
The "junction" is on the silicon die, heat must travel from die to the metal TO220 tab,
and from there get carried off into the air.
    https://media.digikey.com/pdf/Data%20Sheets/Fairchild%20PDFs/LM78T05.pdf
Thermal resistance junction to case", that is from die to the TO220 tab 
is 2.5 degrees C per Watt.  Useful to know if you have a good heatsink for it.

Assume we are supplying the LM78T05 with 12 Volts and drawing 1 Amp at 5 Volts from it.
The LM7805 is then dissipating  (12v-5v) * 1a = 7 Watts,  and if being used without
a heatsink then the tab will be 7*65= 455 degrees C (819 F) above the ambient air temperature.
If the ambient air is 23 degrees C (73 F), then the tab will be around 23+455 = 478 C (892 F).

Actually, it will never get anything like that hot. 
The device goes into thermal shutdown (and the output goes to zero Volts) when it gets too hot,.
Not explicitly stated in the datasheet but I assume somewhere around 125 C (257 F).
(It also has a separate mechanism to limit the maximum current
in case you manage to short the output with a screwdriver.)

So if the device seems to work for a few seconds and then quits, touch the tab with a wet finger. 
If your finger sizzles, then the LM78T05 (or LM7805) is too hot.

I can believe that some devices out there sold for cheap are factory seconds
that don't meet spec, or rejects that don't work at all, or some totally different device
that has been painted over with an "LM78T05" logo.

A test for basic function would be to put a 50 ohm resistor (able to dissipate at least 1/2 Watt)
across the 5 Volt output, so it is only being asked to give  5v/50 = 0.1 Amp.
Then measure the output voltage, see if it really is 5 Volts.

The Fairchild "thermal resistance junction to air" figure of 65C/Watt
is much higher than TI's LM7805 figure of 19C/Watt.
  https://www.sparkfun.com/datasheets/Components/LM7805.pdf
Perhaps due to different assumptions about available air flow?
The "thermal resistance junction to case" figures are fairly close,
2.5C/W for the LM78T05 vs 3.0C/W for the LM7805.
Both are TO220 devices, so the additional thermal resistance from case to air
should also be about the same.

Jerry, KE7ER


On Tue, Sep 17, 2019 at 04:05 AM, Razvan (M0HZH) wrote:

Why would you need another regulator ? The LM7805 is designed for 1A output and the 5" Nextion display + Raduino uses about 500mA (per specs).

You only need a heatsink for the LM7805.

This will be true for any other linear replacement you use (KA78T05 / 78S05 etc) as the dissipated power is the same.

Re: Never used V3 uBitx Board Available

Bill Lamm
 

Did this sell yet?

Re: Nextion with KA78T05

Herman Scheper
 

You are 100% right!! tnx

 

Van: BITX20@groups.io <BITX20@groups.io> Namens Razvan (M0HZH)
Verzonden: dinsdag 17 september 2019 13:06
Aan: BITX20@groups.io
Onderwerp: Re: [BITX20] Nextion with KA78T05

 

Why would you need another regulator ? The LM7805 is designed for 1A output and the 5" Nextion display + Raduino uses about 500mA (per specs).

You only need a heatsink for the LM7805.

This will be true for any other linear replacement you use (KA78T05 / 78S05 etc) as the dissipated power is the same.

Re: Nearly non-existent output power, no mod

Adrian Chadd
 

Hi,

Maybe check the firmware and make sure there's a delay between switching the relays and enabling TX RF?



-a


On Tue, 17 Sep 2019 at 04:15, Ted via Groups.Io <k3rta=yahoo.com@groups.io> wrote:
Clark,

Please forgive my lack of clarity.  I'll try to fashion up something better and make a go of individual relay measurements within a week or so; I'm about to have vertebrae C3-C7 fused and bracketed together, so, any messages after 18 Sept may be Oxycontin-laced  :O   At least I'll have plenty of time.

My chief focus was to answer the concern of whether the relays were switching, i.e. if RF was being sent through the wrong LPF for all but one band, or through more than one filter at once, etc.  I get the impression that the filters are switching in when they 'sposed to, in country parlance, though I'll grant that this fails to say whether good connections are being made in the process.  Again, I'll rig up a better test bench and do a more precise analysis.

I made the measurements using a simple oscilloscope with a cap on the pointy end of the probe.  It picked up some stray RF on adjacent points but for the nature of the test, I let that pass. 

RE: "Bias" resistors, I'd assumed that people who've been around for a while here know that when switching from 3904's to 2222's in the driver stage, it's the 20-ohm [emitter] bias resistors that get doubled up for a net 10-ohms.  My bad.  On the matter of the 2N2222's, I found that transistor mod very worthwhile. 

I'll report back....


Tnx,

Ted
K3RTA

Re: Nextion with KA78T05

Herman Scheper
 

Hello Jerry,

 

I ordered 5 pcs. of them. After it did not work-from the 1th second!! )- i tested 2 of them on a small breadboard. Without load and with a load of a few 100 Ohms.

The did never work. Always output  zero Volt.   Some other members on this forum had the same experience with them.

So never buy this parts form Aliex.

 

Tnx fort he explanation.

 

Rgds,  Herman  PA0BAB

 

 

Van: BITX20@groups.io <BITX20@groups.io> Namens Jerry Gaffke via Groups.Io
Verzonden: dinsdag 17 september 2019 16:13
Aan: BITX20@groups.io
Onderwerp: Re: [BITX20] Nextion with KA78T05

 

Could be that you got bad parts.

Or could be that the part is getting too hot, and going into thermal shutdown.
So you need a better heatsink.


On page 2 of this Fairchild datasheet for the LM78T05, it says that
the "thermal resistance junction to air" is 65 degrees C per Watt.
The "junction" is on the silicon die, heat must travel from die to the metal TO220 tab,
and from there get carried off into the air.
    https://media.digikey.com/pdf/Data%20Sheets/Fairchild%20PDFs/LM78T05.pdf
Thermal resistance junction to case", that is from die to the TO220 tab 
is 2.5 degrees C per Watt.  Useful to know if you have a good heatsink for it.

Assume we are supplying the LM78T05 with 12 Volts and drawing 1 Amp at 5 Volts from it.
The LM7805 is then dissipating  (12v-5v) * 1a = 7 Watts,  and if being used without
a heatsink then the tab will be 7*65= 455 degrees C (819 F) above the ambient air temperature.
If the ambient air is 23 degrees C (73 F), then the tab will be around 23+455 = 478 C (892 F).

Actually, it will never get anything like that hot. 
The device goes into thermal shutdown (and the output goes to zero Volts) when it gets too hot,.
Not explicitly stated in the datasheet but I assume somewhere around 125 C (257 F).
(It also has a separate mechanism to limit the maximum current
in case you manage to short the output with a screwdriver.)

So if the device seems to work for a few seconds and then quits, touch the tab with a wet finger. 
If your finger sizzles, then the LM78T05 (or LM7805) is too hot.

I can believe that some devices out there sold for cheap are factory seconds
that don't meet spec, or rejects that don't work at all, or some totally different device
that has been painted over with an "LM78T05" logo.

A test for basic function would be to put a 50 ohm resistor (able to dissipate at least 1/2 Watt)
across the 5 Volt output, so it is only being asked to give  5v/50 = 0.1 Amp.
Then measure the output voltage, see if it really is 5 Volts.

The Fairchild "thermal resistance junction to air" figure of 65C/Watt
is much higher than TI's LM7805 figure of 19C/Watt.
  https://www.sparkfun.com/datasheets/Components/LM7805.pdf
Perhaps due to different assumptions about available air flow?
The "thermal resistance junction to case" figures are fairly close,
2.5C/W for the LM78T05 vs 3.0C/W for the LM7805.
Both are TO220 devices, so the additional thermal resistance from case to air
should also be about the same.

Jerry, KE7ER


On Tue, Sep 17, 2019 at 04:05 AM, Razvan (M0HZH) wrote:

Why would you need another regulator ? The LM7805 is designed for 1A output and the 5" Nextion display + Raduino uses about 500mA (per specs).

You only need a heatsink for the LM7805.

This will be true for any other linear replacement you use (KA78T05 / 78S05 etc) as the dissipated power is the same.

Re: Nextion with KA78T05

Mark Hatch
 

Perhaps thermal overload, but I was having problems with the 7805 until I replaced it when I put in a 5" Nextion. I did not try the heat sink as it required more cutting to fit then I wanted to try. I also planned on additional boards (stand alone signal analyzer and JACKAL, so a higher rated 7805 mounted on the back panel (using the case as the heat sink) worked well for me. 

Mark

Re: SDR and Ubitx freq not the same.

Bill Lamm
 

I have had the coffee today.. maybe need to keep a white board at my side when I read these posts?!