|
One of the NanoVNA users on Facebook applied too much power to Port 1 of his NanoVNA-F took out three resistors (R36, R39 and R47) in his front end (see attachment). Its a reminder that before you connect your NanoVNA to a device for testing you should ensure the input power rating of the NanoVNA is not exceeded.
So what is the maximum power that can be applied to the input port? Leon Huang, the designer of the NanoVNA-F, replied that it is +20 dBm as shown in the second attachment. All NanoVNA variants using a resistive bridge share the same front end design, so I think the +20 dBm specification is applicable across the board.
Also note in the partial schematic Leon provided that the NanoVNA port input is DC coupled, so if the device you are testing is dc powered ensure the dc is decoupled from the NanoVNA test port. If not, similar damage may occur from DC power exceeding 100 mW.
- Herb
|
As an additional comment on this - I would advise anyone experiencing a burnt front-end bridge to actually replace ALL the resistors in the bridge as they will all have been over-stressed.
Those would be R9,10,11,12,14 on the TX Ch0 or R22,23,24,25 on RX Ch1
At the same time, you can refer to Hugen's latest V3.4 schematic (on github) and update the resistor values to that of the NanoVNA-H
In addition, some forum members have suggested mounting the resistors face down to reduce 'lead length'.
YMMV
toggle quoted message
Show quoted text
On Monday, April 13, 2020, 10:20:17 a.m. GMT-4, hwalker <herbwalker2476@...> wrote:
? ? One of the NanoVNA users on Facebook applied too much power to Port 1 of his NanoVNA-F took out three resistors (R36, R39 and R47) in his front end (see attachment).? Its a reminder that before you connect your NanoVNA to a device for testing you should ensure the input power rating of the NanoVNA is not exceeded.
? So what is the maximum power that can be applied to the input port?? Leon Huang, the designer of the NanoVNA-F, replied that it is +20 dBm as shown in the second attachment.? All NanoVNA variants using a resistive bridge share the same front end design, so I think the +20 dBm specification is applicable across the board.
? Also note in the partial schematic Leon provided that the NanoVNA port input is DC coupled, so if the device you are testing is dc powered ensure the dc is decoupled from the NanoVNA test port.? If not, similar damage may occur from DC power exceeding 100 mW.
- Herb
|
Anyone frying the input needs to pay more attention to what they are doing.
Or maybe learn how to make measyrements.
I'd bet it was many watts and unintentional. That said over 15dbm is too much.
and applying DC across the inputs is really bad practice as then the collective
DC+RF heating is likely to cook things.
Allison
-----------------
No direct email, it goes to bit bucket due address harvesting in groups.IO
|
On Mon, Apr 13, 2020 at 11:11 AM, aparent1/kb1gmx wrote:
Anyone frying the input needs to pay more attention to what they are doing. Or maybe learn how to make measyrements.
I'd bet it was many watts and unintentional.
=======================================
Allison,
The NanoVNA is a new type of measurement instrument for a lot of users. The input power specification and type of input coupling are not commonly published along with other performance specifications, so lack of knowledge rather than attention to detail is a more likely culprit for blown front ends. If it happens a second time then its more likely inattentiveness. I don't think anyone will take you up on your bet about it being unintentional. No one would intentionally fry their front end. Also, some new users may mistakenly assume the NanoVNA is an RF device, and therefore; is ac coupled, thus the dc power warning.
I agree that I wouldn't push the +20 dBm ceiling.
I created the topic to inform those users who were not aware, or needed reminding, about the port power limit specifications. More advanced users need not apply.
- Herb
|
I recently tested every pre-amp in my collection with the nano-VNA. Something to do while we are shut in the house social distancing. Because I did not know how much power the input port of the VNA would handle I made the simplifying assumption that if I nulled the marked gain of the pre-amp I should avoid smoke.
This strategy seemed to work pretty well, as may be seen from the plot of the 38 dB 1090 MHz ADS-B pre-amp, with filters. I used two 20 dB attenuators in series with the output of the pre-amp. It would have been better to also take the maximum permissible input of the pre-amplifiers into consideration, but none of the pre-amps tested are terribly valuable and all survived.
--
Earl, 4Z4TJ
|
Thanks Herb, while I haven't made such an error, I appreciate the reminder. If my circuit under test is not immediately known I add attenuators and back into the signal, much like what one does with a volt meter when a autoDMM isn't available ;-)
|
Measured an amp as well, used a -40dB device in front of the amp.
Good results.
But with preamps, better use a -80dB device, then the attenuated signal of de nanovna is more in line with normally used signal levels for preamps.
Arie PA3A
Op 14-4-2020 om 05:39 schreef Earl:
toggle quoted message
Show quoted text
I recently tested every pre-amp in my collection with the nano-VNA. Something to do while we are shut in the house social distancing. Because I did not know how much power the input port of the VNA would handle I made the simplifying assumption that if I nulled the marked gain of the pre-amp I should avoid smoke.
This strategy seemed to work pretty well, as may be seen from the plot of the 38 dB 1090 MHz ADS-B pre-amp, with filters. I used two 20 dB attenuators in series with the output of the pre-amp. It would have been better to also take the maximum permissible input of the pre-amplifiers into consideration, but none of the pre-amps tested are terribly valuable and all survived.
|
An be aware that for frequencies > 300 MHz the NanoVNA uses harmonics. The fundamental is much stronger and may cause the amplifier to saturate the amplifier. The use of a band pass filter is recommended. You can include it in the calibration to compensate its insertion loss.
Reinier
Op 14-4-2020 om 12:02 schreef Arie Kleingeld PA3A:
toggle quoted message
Show quoted text
Measured an amp as well, used a -40dB device in front of the amp.
Good results.
But with preamps, better use a -80dB device, then the attenuated signal of de nanovna is more in line with normally used signal levels for preamps.
Arie PA3A
Op 14-4-2020 om 05:39 schreef Earl:
I recently tested every pre-amp in my collection with the nano-VNA.? Something to do while we are shut in the house social distancing.?? Because I did not know how much power the input port of the VNA would handle I made the simplifying assumption that if I nulled the marked gain of the pre-amp I should avoid smoke.
This strategy seemed to work pretty well, as may be seen from the plot of the 38 dB 1090 MHz ADS-B pre-amp, with filters. I used two 20 dB attenuators in series with the output of the pre-amp.? It would have been better to also take the maximum permissible input of the pre-amplifiers into consideration, but none of the pre-amps tested are terribly valuable and all survived.
--
IDcircuits - RFID & Electronics
Het Halster 40
6581 JL Malden
The Netherlands
+31 633702492
www.idcircuits.com
r.gerritsen@...
|
Reiner and Arie thanks for your helpful comments.
I have learned something useful for future measurements. I should have mentioned that the preamp under test incorporates two SAW filters 1090 MHz +/- 15 MHz, so, in the plot I attached, the fundamental frequency likely did not have much effect, however it would explain some anomalous results with other amplifiers that I tested.
--
Earl, 4Z4TJ
|
Herb,
Thanks for the information.
I will say that I have used the NanoVNA to demo our ICS preamps at several shows where the unit ran for 6 to 8 hours steady with no issues.
Our preamps have 15 db gain so with the input at -10 db, the output from the preamp to Port 2 was in the +5 db neighborhood.
No complaints so far. Even with the hard life it has had the unit is just as happy as it can be.
Joe - WA7JAW
|
On Tue, Apr 14, 2020 at 10:53 PM, Joe Burkleo wrote:
I will say that I have used the NanoVNA to demo our ICS preamps at several shows where the unit ran for 6 to 8 hours steady with no issues.
Our preamps have 15 db gain so with the input at -10 db, the output from the preamp to Port 2 was in the +5 db neighborhood.
No complaints so far. Even with the hard life it has had the unit is just as happy as it can be.
==============================================================
Joe,
Thanks for the real world use feedback. The +20 dBm port specification is actually pretty robust compared to other high end test equipment I 've used. I believe the receiver port on the HP8753 VNA was specified at 0 dBm, although actual destructive levels was probably more around the +20-30 dBm range.
I've also tested pre-amps without damaging the NanoVNA. I normalize the test set-up by using the same amount of in-line attenuation as the expected gain of the pre-amplifier. I haven't tested my rf power amps (10-20 watts) yet, and thinking I might use couplers for that testing rather than attenuators.
The cautions about fundamentals being higher than the test frequencies when sweeping above 300 MHz apply to the NanoVNA variants. The S-A-A-2 does not use harmonics for high frequency testing so it may be preferable for sweeping active devices. The waveforms output by the ADF4350 and si5351 synthesizers in the S-A-A-2 are not sinusoidal so harmonics are produced, but they are not intentionally used for high frequency operation as in the NanoVNA design.
- Herb
|
On Tue, Apr 14, 2020 at 10:53 PM, Joe Burkleo wrote:
Our preamps have 15 db gain so with the input at -10 db, the output from the
preamp to Port 2 was in the +5 db neighborhood.
Joe, I think some of the latest firmware by Hugen has been increasing the output from the NanoVNA from around -10 dBm to 0 dBm so this will affect how much gain a preamp can have before damaging the input of CH1. Roger
|
At work, where the microwave VNA's cost over $100,000, and the Spectrum Analyzers are around $95,000, we have fellow
engineers cross-check each other before applying power. We use boxes of attenuators and adapters. We fear making a
costly mistake like too much input power. They all have max input power levels ranging from 0 dBm to +30 dBm, depending
on brand.
Switching context over to the NanoVNA; There is a certain pleasure, a smile, and ease of sleeping at night,
knowing that they only cost $50 each. No mistake will ever cost me more than $50. Heck, a speeding
ticket is 10 times that around here :-) Dropping a good glass of red wine can almost cost that much :-)
And, depending how busy you are and what you value your labor rate at, it may be cheaper to keep several
NanoVNA's on the shelf, rather then the time spent soldering/repairing/re-testing one that you burn. :-)
Cheers,
Neal
|
On Wed, Apr 15, 2020 at 02:16 PM, Neal Pollack wrote:
... Switching context over to the NanoVNA; There is a certain pleasure, a smile, and ease of sleeping at night, knowing that they only cost $50 each. No mistake will ever cost me more than $50....
========================================
Neal,
I totally agree. Since buying the NanoVNA, I've learned more in the last year about using a network analyzer then I have in the previous 5 years of working with a HP8753C VNA. Mainly because there are measurements that I can grab and try on the NanoVNA that I would never perform on the HP8753C out of fear of mistakenly making a incorrect connection that would result in costly repairs.
I can also control and collect data from the NanoVNA using open source software that is unavailable for the HP8753C. The GPIB interface module for talking to the HP8753C can set you back the cost of 3 to 4 NanoVNA's. For hobby use, the NanoVNA meets all my requirements.
- Herb
|
Herb, thanks so much for this info. Just the thing I was looking for last evening while using my SAA-2N to test a low priced LNA I found on line (SPF5122Z). Since I was not sure (yesterday), I used the same setup I used when I tested it with my HP8753A and HP85046A. Decreased the output (source) of the NanoVNA to -10dBm, and had two step attenuators (at max 12 dB) in the input and output respectively.
73,
Prem
VA3UMA
|
Earl
Les
Arie Kleingeld PA3A
Roger Need
Neal AB6G
Prem va3uma