Q-meter questions

[regenfreak]

I am looking for a Q-meter to be used for testing homebrew RF, IF and oscillator coils.
PM please for offer.

I have been measuring Q with the ring-down method but it is not accurate with basic scope and function generator.

At the moment I can only predict unload Q with software.

Sing

[Skywave]

I don't have a Q-meter for sale. However, I would recommend you look for an Advance T2 Q-meter. I have one here and is a useful piece of test equipment. The lowest freq. that it will resonate a coil to is 100 kHz. If you do acquire one, you will need the manual: it's available on the www.
Marconi made a Q-meter. I've never used one, but I have read reports that it a bit tricky to use. (And, I might add, in my experience, as an item from Marconi, that is what I would expect).

Al.

[regenfreak]

Thanks for the advice. I don't see them around often.

[Argus25]

By far and away the best Q meter I have seen was a "build your own version" in Electronics Australia, June 1969.

The thing that makes me laugh about it, they called it the "Economy Q Meter" It is far far better than most Q meters and is a wonderful project to build, just as good now as when they published it.

I've attached some scans, obviously not the whole article. But the whole design of it, the mathematical support, operating instructions, the construction article and details are just wonderful.

If you want the whole article, all you have to do is email Silicon Chip (they have this archived) and they will send you the entire article.

[regenfreak]

Thanks the article. Without a Q-meter, it is not the end of the world. I can always rely on design rule of thumb and approx. known values such as graphs from Terman:
https://www.robkalmeijer.nl/techniek...e33/index.html

[G0HZU_JMR]

Back in the 1990s I used the classic Marconi TF1245 Q meter to aid in design for high performance LPF, HPF and BPF filters. This was generally for filters across all bands from <100kHz to a few hundred megahertz. But this meter is big and bulky especially when fitted with both oscillator options and it needed lots of maintenance to keep it working on all its ranges. So I can't recommend it in 2019. Probably the best Q meter in those days was the HP4342A but these are rare and expensive unless you get lucky.

If you don't manage to find/buy a Q meter then there are various (fairly simple) methods you can use to measure inductor Q across the HF bands and up into VHF.

I've not used a Q meter in probably 15-20 years because I don't think they are needed for general RF design in 2019. Because of the abundance of accurate test generators and detectors these days it is possible to lash something up (to measure Q) in a basic test jig that will be more trustworthy than a 40-50 year old Q meter that will have seen a few repairs and bodges during its lifetime.

So I'd argue these old Q meters are a waste of space unless you need to measure the Q of lots of inductors in a short time. The only exception is the HP 4342A Q meter. I wouldn't say no to one of those but they are rare and expensive!

[regenfreak]

Hi Jeremy
Thanks for the advice. I agree. It is one of those things that it is nice to have but there are many ways of working around it.

Funny enough, the only reason that I had the "desire" to buy a Q-meter was that I was reading the manual for the HP4342 last week. Someone in another forum had suggested me to download the manual and look at the method used in HP4342 instructions of measuring the stray capacitance of an inductor . But the method had turned out to not useful. But I must say the HP4342 manual has one or two interesting graphs and I was intrigued by them.
There are a few online unloaded Q calculators, like this modern one using "helical waveguide mode" method (it seems very advanced but is not very good for working stray capacitance):

http://hamwaves.com/inductance/en/index.html#input

[Skywave]

Many years ago, a design for a home-build Q-meter appeared in an issue of Practical Wireless magazine. Several years later, I bought a second-hand one. Briefly stated, it was almost useless. So I suggest that if you do see one offered, walk away!

Al.

[G0HZU_JMR]

Originally Posted by regenfreak

Someone in another forum had suggested me to download the manual and look at the method used in HP4342 instructions of measuring the stray capacitance of an inductor . But the method had turned out to not useful. But I must say the HP4342 manual has one or two interesting graphs and I was intrigued by them.
There are a few online unloaded Q calculators, like this modern one using "helical waveguide mode" method (it seems very advanced but is not very good for working stray capacitance):

http://hamwaves.com/inductance/en/index.html#input

Yes, I've already read quite a bit of the stuff produced by David Knight and the other people in your hamwaves link. They seem very focussed on predicting 'where' the resonances will occur and then producing an equivalent 'capacitance' to produce this resonance. Or at least that was what I concluded the last time I read their stuff. I think David Knight has been doing research work into this stuff for many years.

My experience of advanced inductor modelling dates back to the early 1990s at my place of work. We have to model inductors out to frequencies several times higher than the first resonance mode because we do a lot of ultra wide bandwidth design. Our goal was to produce real working (complex transmission line based) models rather than to try and produce equations to predict where resonances occur. However, these days we mainly use 2 port VNA models for SMD inductors as this is quick, convenient and is usually adequate for SMD inductors that have lowish Q. But our old transmission line based models from the 1990s were very powerful and much better than the classic/flawed inductor models based on lumped components.

[G0HZU_JMR]

Coilcraft offer some info on how to model SMD inductors using a crude transmission line model here:

https://www.coilcraft.com/pdfs/spice_132.pdf

This is better than a lumped model but still fairly hopeless as a wideband model (out to several GHz) as it won't model the higher resonance modes correctly.

... but none of this will help you buy that Q meter. Someone on here may have one for sale. If it's an HP4342A I'll out bid you for it (just kidding )

[regenfreak]

Thanks. I don't know much about Ham radio but every time I look for any information I ended up reading articles written by Ham guys. I am only interested in unloaded Q prediction up 10 MHz for MW tank/oscillator radio and tesla coils design. I would stay away from the "black magic" of transmission line theory up to GHz. The maths of the VHF stuff boggles my mind.

It appear that the HP4342 is "an object of desire" then. It also seems that the owners of the HP4342 (guys from another forum) do not normally understand the logic of the testing steps outlined in the manual. Last week I was trying to figure out why certain test have to be done in specific steps and the HP4342 manual does not really explain the rational behind those testing procedures.

Other less known Q meters would require lots of skills to restore and maybe it would be difficult to use correctly without the user manuals.

[G0HZU_JMR]

Wow, I'm not so sure about testing Tesla coils for Qu. I keep away from scary stuff like Tesla coil/machines so I have no idea what spec of Q meter you would need. Most of the RF stuff I do these days is SMD but in the 1990s I did make a lot of coils for filters across LF through VHF and these were fairly large coils with lots of turns. I often used Litz wire for the LF stuff with various RM cores from Siemens to achieve a Qu up in the 400-500 range.

Most Q meters I've seen operate on the principle that Q can be approximated by E/e. Whilst this is an elegant way to do it, it does make it hard to realise the source and detector requirements if you want to test coils that have a low ESR or high Rp at the test frequency. That's when the decent Q meters leave the cheap and basic Q meters behind. It's also why I prefer to use other means to measure Q these days because I feel more confident in the results.

But for simple and typical MW coils that have moderate Qu I think you would be OK with a fairly old and basic Q meter. I can't comment on your Tesla coil requirements though!

[Philips210]

Originally Posted by Argus25

By far and away the best Q meter I have seen was a "build your own version" in Electronics Australia, June 1969.

Thank you Hugo, that looks an excellent project to build, very useful for checking home made RF coils. When time permits I'll be building this project.

Regards
Symon

[Ed_Dinning]

Hi Guys, the best and simplist Q meter for the OP's application is the Advance T2.
It is simple to use and is fitted with calibrated variable C and goes from 50KHz to 100MHz.
It will allow Q and delta q measurements, as well as dynamic R and self cap of the test coil.
It has a small platform that allows accurate measurements. A rats nest of wires will not give accurate Q figures.
Parts to replace are black hunts caps that are not in the measuring circuit and valves (2) are easy to obtain types. Used to be quite common on ebay but not seen one for a while.
Use with a Tesla coil would only be at very low power as the coil would only be energised form the meter.


Ed

[regenfreak]

Originally Posted by G0HZU_JMR

Wow, I'm not so sure about testing Tesla coils for Qu. I keep away from scary stuff like Tesla coil/machines so I have no idea what spec of Q meter you would need. Most of the RF stuff I do these days is SMD but in the 1990s I did make a lot of coils for filters across LF through VHF and these were fairly large coils with lots of turns. I often used Litz wire for the LF stuff with various RM cores from Siemens to achieve a Qu up in the 400-500 range.

But for simple and typical MW coils that have moderate Qu I think you would be OK with a fairly old and basic Q meter. I can't comment on your Tesla coil requirements though!

This is exactly the reason I have stopped designing and building tesla coils.I was building both spark gap and double resonance tesla coils. My high-power tesla coil scares the living hell out of me. It was a health and safety hazard. I had to wear gas mask and ear muffs during operation..... It is important to predict the Q and coupling coefficient k before building a tesla coil. This can be done with a very accurate free online software called Javatc. The measurement of Q for tesla coils can be done in the same way as radio RF coils.

Just out of historical interest in vintage technology, I made a few spiderweb RF coils with 100 threads Litz wire, I think the Qu of spiderweb coils can go over the maximum limit of some Q meters; probably Qu reaches over 400. In that case software prediction or measurement methods using scope and signal generator are the only way forward.

T2A looks beefy and solid.

The maximum Q for HP4342A is 300 (up to 1000 with certain supplemental inductor combo)
The maximum Q for T2A is 400

[Radio Wrangler]

Originally Posted by regenfreak

It appear that the HP4342 is "an object of desire" then. It also seems that the owners of the HP4342 (guys from another forum) do not normally understand the logic of the testing steps outlined in the manual. Last week I was trying to figure out why certain test have to be done in specific steps and the HP4342 manual does not really explain the rational behind those testing procedures.

Other less known Q meters would require lots of skills to restore and maybe it would be difficult to use correctly without the user manuals.

I used an HP4342A for many years while I was developing filters and low noise oscillators. It was a replacement for an old Boonton 260 (I think the model number was) It could certainly do a better job on high Q coils than you could do with a VNA but even so, you could run into its limitations. I have a feeling it was one of the first products of HP's Japanese subsidiary with Yokogawa. Search the HP Journal , there is likely to be an article on it explaining all sorts of things, look also for HP applications notes.

There was also a Boonton Rx meter and of course, General radio's RF bridge. But give me a 4342A any day

David

[G0HZU_JMR]

If there is interest in discussing this stuff in detail then maybe it would be better to do it on a dedicated thread in the Test Gear section. I would certainly contribute to it. I've measured Q in lots of ways over the years and I abandoned the old Q meters many years ago. We threw our Marconi TF1245 away some time around 2000-2005 as no one had used it in ages. The TF1246/7 oscillator coils tend to fail on certain ranges (they can be fixed) and we had trouble every year we sent the complete 1245/6/7 assembly for 'calibration'.

Typical issues were to do with the cal house wanting to engrave a new capacitance dial (at immense cost) every other year or so and they also bleated every year about the detector diode (valve) failing their tests. They wanted to import the diode valve from the US at a crazy cost and they wouldn't guarantee the new diode would be any better than the old one. So every year we asked for a restricted calibration rather than spend money on it. That's why I would urge caution with this meter. If you find one it's probably had several sets of fingers in it to fix the usual failures and it will be unlikely to be in spec any more.

However, I have fond memories of using it because it can be very rewarding seeing that Q needle climb just a little bit higher than it did with a previous coil design! I was chasing coil Q of >400 down at frequencies below 1MHz in the early days and I spent a LOT of time playing with pot cores and various types of Litz wire. I don't think the meter was very accurate when used like this but I was looking for increases rather than trust the Q dial.

[G0HZU_JMR]

Originally Posted by regenfreak

Just out of historical interest in vintage technology, I made a few spiderweb RF coils with 100 threads Litz wire, I think the Qu of spiderweb coils can go over the maximum limit of some Q meters; probably Qu reaches over 400. In that case software prediction or measurement methods using scope and signal generator are the only way forward.

Assuming you are doing this stuff down in the 150-1600kHz region, I'll have a play with my Analog Discovery II over the next few days. It should be possible to lash a jig together (to measure a typical LW or MW coil with a Q of several hundred) that will work quite well down here. I think the price of the AD2 has been slowly creeping up but I got mine a couple of years ago for something like £250 with a few accessories. This is probably a lot more than the little Advance T2 Q meter will cost but the AD2 is a very versatile tool. I'm impressed with mine but I really don't use it enough as most of the stuff I do is up at UHF and beyond.

[Ed_Dinning]

Hi, it appears that the OP was looking for comparative measurements, so that absolute accuracy is not required and frequency can easily be checked with a cheap frequency meter.
If the Q goes beyond the 400, then it is a simple matter to add a fixed resistor across the coil for measurement (that are again comparative).

I did read an early WW article (20's I think) about special coils that had a peak Q of 5000, but were quite only stable for short periods of time. I'm not now sure what construction methods were used.

Ed

[saddlestone-man]

Hello

The Electronics Australia, June 1969 Q-meter design looks very useful, and easily reproducible today. The full article covers six pages and a scan of this issue of the magazine can be found at: https://www.americanradiohistory.com...stralia_AU.htm

best regards ... Stef