|
Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members. |
|
Thread Tools |
27th Dec 2017, 5:56 pm | #1 |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Homebrew FET voltmeter mark 2
Due to an unfortunate accident involving over-excited children not doing as they were told, the FET voltmeter I built in April leapt from the table to it's demise. The meter movement was virtually destroyed by the drop. Some of the insides, due to the precarious battery placement and sheer mass of 8xAA batteries were also pretty heavily damaged.
RIP: https://www.vintage-radio.net/forum/...d.php?p=942748 So about two weeks ago I decided to build a successor. This one is designed from scratch and has a number of new, improved features: 1. No zeroing required. 2. Better linearity. 3. A power LED (as I kept leaving the old one on). 4. Better quality, larger meter movement with mirror. 5. The input terminals are 0.75" apart to match "banana plug pairs" 6. Different orientation makes it less likely to fall over. 7. Single 9v battery. 8. Chicken head knob so you can actually see what range it is in. 9. Uses hand selected standard low quality parts instead of precision resistors. 10. Better PCB design methodology than historical ones. 11. Better ranges. Unfortunately I don't have a schematic other than a very early one to hand, but I'll describe the circuit fully soon ... The words below describe a number of corrections to the schematic above. Please bear those in mind! The input full-scale deflection is 300mV. This is converted to a current using a CA3140E based transconductance amplifier (voltage to current). The ratio is set at just under 50uA per 300mv and adjusted "up" using a trimmer. The drive transistor is a 2n3904 (any NPN should be fine here). Everything is absolutely 0V referenced so no bridge is required. There is a 4.3V zener in front of the input which provides both forward over-voltage and reverse over-voltage protection for the op-amp. Preceding this, there is a 1M/10nF low-pass filter. And before that, there is a divider consisting of hand selected 5% 0.25W resistors (nothing fancy). The resistor values in the divider are: 8M, 1M, 800K, 100K, 80K, 20K totalling 10M which sets the input impedance. This gives 300mV, 1.5V, 3V, 15V, 30V, 150V ranges. These were chosen by convenience and that they are slightly larger than common voltages like 13.8V so reasonably accurate measurements can be made. This is all chucked in a simple ABS box (bitsbox), rotary switch (recovered from old meter), meter movement (Rapid) and some banana sockets (recovered), knob (bitsbox), 9v holder (bitsbox), 1/0.6 wire (tayda). PCB was hand drawn with Staedtler Lumocolor blue marker, drilled with a Dremel and etched in the sink with ferric chloride. By pure luck, the LED hangs off the reverse side of the board and lines up nicely with the front panel. Total expenditure was around £23 to complete and most of that was the meter movement. To calibrate, short input terminals, power up and adjust meter screw to zero. Then connect to 15.0V supply (I use my calibrated PL310 for this) and change the trimmer inside for full scale reading. Need to do a meter face for it with 15 and 3 as full scale and label it all yet but otherwise finished. Could do with some 2p pieces glued into the bottom of the case to give it a little more weight too. Merry Christmas and a happy new year to all. Edit: I have abandoned the previously mentioned 12AU7 based VTVM in favour of the above plus the final acquisition of a couple of very much limping Heath V7's really put me off the damn things forever. Last edited by MrBungle; 27th Dec 2017 at 6:01 pm. |
27th Dec 2017, 6:06 pm | #2 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,924
|
Re: Homebrew FET voltmeter mark 2
Thanks for that; I'll file it. Long ago when Al and myself were pursuing RF voltmeters, I built one circuit using the 3140, but you have added some useful tweaks.
B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
27th Dec 2017, 6:07 pm | #3 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: Homebrew FET voltmeter mark 2
There's an error due to the 2N3904 base current, but it ought to be trivial.
The problem I found with current driven meters is the lack of damping. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
27th Dec 2017, 6:08 pm | #4 |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Re: Homebrew FET voltmeter mark 2
It's a glorious opamp, if a little expensive. RF voltmeter/load is next task actually. If you have any knowledge to share I'd appreciate it
|
27th Dec 2017, 6:11 pm | #5 | |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Re: Homebrew FET voltmeter mark 2
Quote:
I found the same thing with the damping hence the high time constant filter on the front end. Seems nice and smooth now. |
|
27th Dec 2017, 6:51 pm | #6 | |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,924
|
Re: Homebrew FET voltmeter mark 2
Quote:
Investigation of the commercial instruments made by Marconi and by Boonton revealed that both use double diode probes with nothing in front of the diodes, but what kind of diodes are they? No one seems to know. One internet site has a bunch of guys taking Boonton probes to pieces. I have some ancient Ge point-contact diodes and they did not seem adequate (for a sensitivity of ~1mV first indication). I think some Schotkys were tried. I eventually gave up; not sure whether Al did better. Always ready to take another look while the external temperature is <10°C B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
|
27th Dec 2017, 7:17 pm | #7 |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Re: Homebrew FET voltmeter mark 2
I will look back and read through the thread.
I was looking at AD8307 solutions for power measurement instead of voltage as such. Seems like it’s easier to get from A-B. I have been playing with those ICs with my SA but I’ve blown up two now. They’re pretty fragile. |
27th Dec 2017, 7:30 pm | #8 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: Homebrew FET voltmeter mark 2
I wouldn't say no-one.
There are ultra-low barrier diodes operated in their square law region at the heart of many RF power sensors. 4v reverse rating, also used in diode ring mixers. HP made their own, sold under the HPA name, then spun out as Avago, and recently re-named Broadcom. Broadcom have just announced the obsoletion of the whole lot along with all the mmic amps, PIN diodes etc. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
27th Dec 2017, 9:50 pm | #9 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,924
|
Re: Homebrew FET voltmeter mark 2
OK, no one in the 'hobby radio' world seems to know what those diodes are. If we did, maybe we could get that young American guy who is making chips in his garage to do some?
Re the AD chips, I think they are configured around a 50R input impedance, whereas the thread was targeted at >500k. B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
27th Dec 2017, 10:19 pm | #10 |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Re: Homebrew FET voltmeter mark 2
Yes they are power meters as opposed to voltmeters.
I suppose, depending on your absolute requirements, you could use a FET buffer, then use an active rectifier. Bob Pease managed 90MHz with old TH parts for an active probe. Tek get a reasonable bandwidth out of some of their old differential probes (matched 2n4416’s) |
27th Dec 2017, 10:26 pm | #11 |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,924
|
Re: Homebrew FET voltmeter mark 2
Maybe stick Skywave's wideband FET amp in front of an AD8307 chip? I bought some of those, 10 for £2.40 from China, and have yet to try them out!
B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
27th Dec 2017, 10:28 pm | #12 |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Re: Homebrew FET voltmeter mark 2
That’s an idea worth trying.
Worth a gamble on the Chinese sourced ones. The ones I’m buying from RS are about £10 each! |
27th Dec 2017, 10:32 pm | #13 | |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: Homebrew FET voltmeter mark 2
Quote:
But there's many a slip 'twixt cup and lip, so don't hold your breath! Al. |
|
27th Dec 2017, 10:42 pm | #14 |
Dekatron
Join Date: Jun 2016
Location: London, UK.
Posts: 3,687
|
Re: Homebrew FET voltmeter mark 2
I have my fingers crossed for you
|
27th Dec 2017, 11:08 pm | #15 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: Homebrew FET voltmeter mark 2
Thank you, Mr. B. On many an occasion, when working on said aforementioned project, I have had many things crossed - including eyes - which doesn't really aid construction!
Hopefully, time and persistence will bear fruit. Al. |
28th Dec 2017, 12:10 am | #16 | |
Dekatron
Join Date: Apr 2009
Location: Oxfordshire, UK.
Posts: 4,924
|
Re: Homebrew FET voltmeter mark 2
Quote:
B
__________________
Saturn V had 6 million pounds of fuel. It would take thirty thousand strong men to lift it an inch. |
|
28th Dec 2017, 6:11 am | #17 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: Homebrew FET voltmeter mark 2
Infineon BAT15 is one of the low barrier detector/mixer diodes. This sort of thing allows RF power detection down to -70dBm, but you have to be very careful because of the 4v max reverse rating.
That -70dBm figure is the full scale of the most sensitive range of some power meters using this class of diode. No amplification and no DC bias tricks are used, but there is a lot of temperature compensation! Some RF power heads can give FSD from -70 to +20dBm, but they switch diode detectors or change from square law to linear law part way through the range. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |