Re: [funwithtransistors] Please Help...W7EL QRP Wattmeter


Jan 28, 2013

 


----------------------------

#1222 Jan 28, 2013

Hi all,



I know this project doesn't have a single tube in it. Since it's going

to be used with tube transmitters using tubes like 6CL6, 6V6, etc, I

decided that I'd post it for the FWTers to see as well. I suppose

there's no reason a directional coupler couldn't be used with a tube

voltmeter.



www.n0ksf.com/SWRmeter/swrmeter.htm



Since the link to the original QST article is on the "members only"

part of ARRL, the above link leads to a schematic and a build of a guy

who made it. There are some parts that I have or will have to

substitute. Attached are pictures of my build.



The meter movement will be one of the 1mA movements I have floating

around somewhere. I will have to substitute the CA3160 chip used in

the diode compensation buffer but it seems that an op-amp like the

TLE2141 will work. I couldn't get FT-37-72 cores for the coupler so I

bought FT-37-61s. Apparently their AL value is about 10% of that of

the 72 cores. I may be able to calibrate around this. I hope.



I hand matched the inductors wound on the toroids and the diodes used

on the couplers. Now all I need to do is get a small piece of metal

and make the meter circuit. The actual meter driver can be made from

any old op-amp but I think a low offset one will work well.



Ed



----------------------------

#1347 May 15, 2013

n0ksf.com/SWRmeter/swrmeter.htm



Hello FWQers,



I am building the "Simple and Accurate QRP Wattmeter" designed by Roy

Lewallen W7EL. I tested it tonight and was able to get a calibration

on. I was sort of successful in the calibration but I ran into some

problems on the lowest (100 mW) range and the meter shows undesirable

operation altogether. The above link shows the schematic. Here's the

issue:



I am trying two different types of op amps for the meter because I

have neither an LM358 or a CA3260. My DIP 8 sockets are wired for the

standard dual op amp chip. I have played with the op amp using both

TL072s and some LM4558s I got in the mail today (after waiting 6 weeks

for them D: ). Here is the normal operation of the meter...the dial

R6 is the "cal" knob. To read power you simply set the pot fully

clockwise. To read SWR you adjust the pot for the meter to read full

scale and then flip the switch to REV.



Here's my problem. You would think that, with no input, the meter

would read zero. If I turn the knob all the way down, the meter pegs

past full scale. If I inch it up, it will eventually dip at a low

value. The value and the place on the dial where the needle dips

depends on the range...10w is the lowest and 100 mW is the highest. If

I continue moving the knob clockwise, the meter keeps increasing.



I did some investigating within the chassis and found this out...the

inverting inputs on both op amps are tied to ground. One would think

that, with no DC input on the non-inverting input, the output would be

very low and close to zero. That is evidently why Roy chose the

CA3160. The op amps (both of them) are latching to near the positive

rail (battery voltage 8.03, positive latch voltage 7.33) when the

non-inverting input pin goes below about 0.78 volts. The first op amp

is doing this because there is no DC input from the directional

coupler. This means that the resistor R6 gets a constant 6.8 volts

across it pretty much all the time. When I crank it down so that it

puts less than 0.8 volts into the input pin of U2, the meter pegs

because the op amp can't handle the low input voltage. When I kick it

up past that, it behaves. Because the diode compensation op amp is

misbehaving, the meter won't function as it should.



So, I must know what to do now. Is this a simple matter of proper chip

selection or is there something else I can do?



Ed AE7TE



----------------------------

#1348 May 15, 2013

Neither the TL072 nor the LM4558 have an input common mode range or

output range which extends to the negative supply which in this case

is going to be a problem under certain conditions. The design could

be changed to fix this by biasing the circuit ground up by a couple of

volts above the negative terminal of the battery.



I wonder why the design did not use the CA3260 for both operational

amplifiers. There are lots of precision CMOS operational amplifies

which would work great as a substitute for both the CA3260 and LM358

in this design like the LMC6062.



On Tue, 14 May 2013 22:22:54 -0700, J Ed jedwardsat1@...>

wrote:

>n0ksf.com/SWRmeter/swrmeter.htm

>

>Hello FWQers,

>

>I am building the "Simple and Accurate QRP Wattmeter" designed by Roy

>Lewallen W7EL. I tested it tonight and was able to get a calibration

>on. I was sort of successful in the calibration but I ran into some

>problems on the lowest (100 mW) range and the meter shows undesirable

>operation altogether. The above link shows the schematic. Here's the

>issue:

>

>I am trying two different types of op amps for the meter because I

>have neither an LM358 or a CA3260. My DIP 8 sockets are wired for the

>standard dual op amp chip. I have played with the op amp using both

>TL072s and some LM4558s I got in the mail today (after waiting 6 weeks

>for them D: ). Here is the normal operation of the meter...the dial

>R6 is the "cal" knob. To read power you simply set the pot fully

>clockwise. To read SWR you adjust the pot for the meter to read full

>scale and then flip the switch to REV.

>

>Here's my problem. You would think that, with no input, the meter

>would read zero. If I turn the knob all the way down, the meter pegs

>past full scale. If I inch it up, it will eventually dip at a low

>value. The value and the place on the dial where the needle dips

>depends on the range...10w is the lowest and 100 mW is the highest. If

>I continue moving the knob clockwise, the meter keeps increasing.

>

>I did some investigating within the chassis and found this out...the

>inverting inputs on both op amps are tied to ground. One would think

>that, with no DC input on the non-inverting input, the output would be

>very low and close to zero. That is evidently why Roy chose the

>CA3160. The op amps (both of them) are latching to near the positive

>rail (battery voltage 8.03, positive latch voltage 7.33) when the

>non-inverting input pin goes below about 0.78 volts. The first op amp

>is doing this because there is no DC input from the directional

>coupler. This means that the resistor R6 gets a constant 6.8 volts

>across it pretty much all the time. When I crank it down so that it

>puts less than 0.8 volts into the input pin of U2, the meter pegs

>because the op amp can't handle the low input voltage. When I kick it

>up past that, it behaves. Because the diode compensation op amp is

>misbehaving, the meter won't function as it should.

>

>So, I must know what to do now. Is this a simple matter of proper chip

>selection or is there something else I can do?

>

>Ed AE7TE







----------------------------

#1349 May 15, 2013

I played around with it and ended up trying a TLC272CP which stopped the crazy. The article requires the use of a voltmeter to cal the device. Soon i will use a lower power transmitter to test the meter on a scope and align it for a more accurate readout. Next will be to redraw the meter face for a log scale and press it into service.EdOn May 14, 2013 11:38 PM, "David" davidwhess@...> wrote:

.Neither the TL072 nor the LM4558 have an input common mode range or

output range which extends to the negative supply which in this case

is going to be a problem under certain conditions. The design could

be changed to fix this by biasing the circuit ground up by a couple of

volts above the negative terminal of the battery.



I wonder why the design did not use the CA3260 for both operational

amplifiers. There are lots of precision CMOS operational amplifies

which would work great as a substitute for both the CA3260 and LM358

in this design like the LMC6062.



On Tue, 14 May 2013 22:22:54 -0700, J Ed jedwardsat1@...>

wrote:



>n0ksf.com/SWRmeter/swrmeter.htm

>

>Hello FWQers,

>

>I am building the "Simple and Accurate QRP Wattmeter" designed by Roy

>Lewallen W7EL. I tested it tonight and was able to get a calibration

>on. I was sort of successful in the calibration but I ran into some

>problems on the lowest (100 mW) range and the meter shows undesirable

>operation altogether. The above link shows the schematic. Here's the

>issue:

>

>I am trying two different types of op amps for the meter because I

>have neither an LM358 or a CA3260. My DIP 8 sockets are wired for the

>standard dual op amp chip. I have played with the op amp using both

>TL072s and some LM4558s I got in the mail today (after waiting 6 weeks

>for them D: ). Here is the normal operation of the meter...the dial

>R6 is the "cal" knob. To read power you simply set the pot fully

>clockwise. To read SWR you adjust the pot for the meter to read full

>scale and then flip the switch to REV.

>

>Here's my problem. You would think that, with no input, the meter

>would read zero. If I turn the knob all the way down, the meter pegs

>past full scale. If I inch it up, it will eventually dip at a low

>value. The value and the place on the dial where the needle dips

>depends on the range...10w is the lowest and 100 mW is the highest. If

>I continue moving the knob clockwise, the meter keeps increasing.

>

>I did some investigating within the chassis and found this out...the

>inverting inputs on both op amps are tied to ground. One would think

>that, with no DC input on the non-inverting input, the output would be

>very low and close to zero. That is evidently why Roy chose the

>CA3160. The op amps (both of them) are latching to near the positive

>rail (battery voltage 8.03, positive latch voltage 7.33) when the

>non-inverting input pin goes below about 0.78 volts. The first op amp

>is doing this because there is no DC input from the directional

>coupler. This means that the resistor R6 gets a constant 6.8 volts

>across it pretty much all the time. When I crank it down so that it

>puts less than 0.8 volts into the input pin of U2, the meter pegs

>because the op amp can't handle the low input voltage. When I kick it

>up past that, it behaves. Because the diode compensation op amp is

>misbehaving, the meter won't function as it should.

>

>So, I must know what to do now. Is this a simple matter of proper chip

>selection or is there something else I can do?

>

>Ed AE7TE



----------------------------

#1350 May 15, 2013

The design, made in 1990, called for a CA3160 and an LM358. I could have done it with a single 272 if i had known what would fix the issue.EdOn May 15, 2013 12:19 AM, "J Ed" jedwardsat1@...> wrote:I played around with it and ended up trying a TLC272CP which stopped the crazy. The article requires the use of a voltmeter to cal the device. Soon i will use a lower power transmitter to test the meter on a scope and align it for a more accurate readout. Next will be to redraw the meter face for a log scale and press it into service.EdOn May 14, 2013 11:38 PM, "David" davidwhess@...> wrote:

.Neither the TL072 nor the LM4558 have an input common mode range or

output range which extends to the negative supply which in this case

is going to be a problem under certain conditions. The design could

be changed to fix this by biasing the circuit ground up by a couple of

volts above the negative terminal of the battery.



I wonder why the design did not use the CA3260 for both operational

amplifiers. There are lots of precision CMOS operational amplifies

which would work great as a substitute for both the CA3260 and LM358

in this design like the LMC6062.



On Tue, 14 May 2013 22:22:54 -0700, J Ed jedwardsat1@...>

wrote:



>n0ksf.com/SWRmeter/swrmeter.htm

>

>Hello FWQers,

>

>I am building the "Simple and Accurate QRP Wattmeter" designed by Roy

>Lewallen W7EL. I tested it tonight and was able to get a calibration

>on. I was sort of successful in the calibration but I ran into some

>problems on the lowest (100 mW) range and the meter shows undesirable

>operation altogether. The above link shows the schematic. Here's the

>issue:

>

>I am trying two different types of op amps for the meter because I

>have neither an LM358 or a CA3260. My DIP 8 sockets are wired for the

>standard dual op amp chip. I have played with the op amp using both

>TL072s and some LM4558s I got in the mail today (after waiting 6 weeks

>for them D: ). Here is the normal operation of the meter...the dial

>R6 is the "cal" knob. To read power you simply set the pot fully

>clockwise. To read SWR you adjust the pot for the meter to read full

>scale and then flip the switch to REV.

>

>Here's my problem. You would think that, with no input, the meter

>would read zero. If I turn the knob all the way down, the meter pegs

>past full scale. If I inch it up, it will eventually dip at a low

>value. The value and the place on the dial where the needle dips

>depends on the range...10w is the lowest and 100 mW is the highest. If

>I continue moving the knob clockwise, the meter keeps increasing.

>

>I did some investigating within the chassis and found this out...the

>inverting inputs on both op amps are tied to ground. One would think

>that, with no DC input on the non-inverting input, the output would be

>very low and close to zero. That is evidently why Roy chose the

>CA3160. The op amps (both of them) are latching to near the positive

>rail (battery voltage 8.03, positive latch voltage 7.33) when the

>non-inverting input pin goes below about 0.78 volts. The first op amp

>is doing this because there is no DC input from the directional

>coupler. This means that the resistor R6 gets a constant 6.8 volts

>across it pretty much all the time. When I crank it down so that it

>puts less than 0.8 volts into the input pin of U2, the meter pegs

>because the op amp can't handle the low input voltage. When I kick it

>up past that, it behaves. Because the diode compensation op amp is

>misbehaving, the meter won't function as it should.

>

>So, I must know what to do now. Is this a simple matter of proper chip

>selection or is there something else I can do?

>

>Ed AE7TE







----------------------------

#1351 May 15, 2013

The TLC272 is another single supply CMOS operational amplifier that

should work fine.



On Wed, 15 May 2013 00:19:15 -0700, J Ed jedwardsat1@...>

wrote:

>I played around with it and ended up trying a TLC272CP which stopped the

>crazy. The article requires the use of a voltmeter to cal the device. Soon

>i will use a lower power transmitter to test the meter on a scope and align

>it for a more accurate readout. Next will be to redraw the meter face for a

>log scale and press it into service.

>

>Ed

>On May 14, 2013 11:38 PM, "David" davidwhess@...> wrote:

>

>> **

>>

>> Neither the TL072 nor the LM4558 have an input common mode range or

>> output range which extends to the negative supply which in this case

>> is going to be a problem under certain conditions. The design could

>> be changed to fix this by biasing the circuit ground up by a couple of

>> volts above the negative terminal of the battery.

>>

>> I wonder why the design did not use the CA3260 for both operational

>> amplifiers. There are lots of precision CMOS operational amplifies

>> which would work great as a substitute for both the CA3260 and LM358

>> in this design like the LMC6062.



----------------------------

#1352 May 16, 2013

Many thanks. I want to express gratitude to TI for sending me samples. I'm going to use them to build some more metrology circuits. I'm sure the RF guys here can give me some ideas for a simple homebrew project.

I do intend to submit media on my RF meter. Some people have been paying attention to it in hopes that they can build the same. The possible substitution of the TLC272 for the CA3260 makes possible this highly useful meter. What I'd like to say is that the industry terminology for such an op amp is "rail-to-rail" op amp. Usually these are compatible with single supplies and can operate right down to the negative rail. Very useful for circuits that need to read zero with no input.

I'd like to make a noise meter that can sense right down to the noise floor of -124 dBm (thanks for that figure, Chuck). That way, I can measure the absolute smallest signals and figure out how to make super-quiet preamps. After I've hit the moon a couple of times, I might take what I've learned and make some audio amps that make audiophiles return their Mexican food rentals.

Ed

On Wed, May 15, 2013 at 12:27 AM, David davidwhess@...> wrote:

.The TLC272 is another single supply CMOS operational amplifier that

should work fine.



On Wed, 15 May 2013 00:19:15 -0700, J Ed jedwardsat1@...>

wrote:



>I played around with it and ended up trying a TLC272CP which stopped the

>crazy. The article requires the use of a voltmeter to cal the device. Soon

>i will use a lower power transmitter to test the meter on a scope and align

>it for a more accurate readout. Next will be to redraw the meter face for a

>log scale and press it into service.

>

>Ed

>On May 14, 2013 11:38 PM, "David" davidwhess@...> wrote:

>

>> **

>>

>> Neither the TL072 nor the LM4558 have an input common mode range or

>> output range which extends to the negative supply which in this case

>> is going to be a problem under certain conditions. The design could

>> be changed to fix this by biasing the circuit ground up by a couple of

>> volts above the negative terminal of the battery.

>>

>> I wonder why the design did not use the CA3260 for both operational

>> amplifiers. There are lots of precision CMOS operational amplifies

>> which would work great as a substitute for both the CA3260 and LM358

>> in this design like the LMC6062.



----------------------------

#1354 May 17, 2013

Hi all,

I performed some measurements today using my QRP SWR Wattmeter. The transmitter was my FT101EE and I used a scope to measure the peak voltage present across a 50w dummy load. I went at it with the following calculations for peak voltage vs. power:

10 watts into 50 ohms is 31.62 volts peak1 watt into 50 ohms is 10 volts peak100 milliwatts into 50 ohms is 3.16 volts peak

First of all, I miswired the FWD-REV power switch so REV gives the accurate reading for power forward. I just need to swap some wires.

Second, I noticed that a full reading on each scale corresponded to a different peak RF voltage depending on the frequency out. For example, on the 10w range, the meter read full scale when the RF voltage was 40.5 volts when I was on the 40 meter band. On the 20 meter band, the same meter reading happened for a peak voltage of 43 volts. On 10 meters, it took 51 volts peak to get the same meter reading.

This tells me that the meter is less sensitive with increasing frequency. Is there anything I can do about it?

Ed



This Site's Privacy Policy
Google's privacy policies

S
e
n
i
o
r
T
u
b
e
.
o
r
g